Research Problems, Research Questions, and Hypotheses

Research Problems, Research Questions, and Hypotheses
Overview of Research Problems
Studies begin, much like evidence-based practice (EBP) efforts, with a problem that needs to be solved or a question that needs to be answered. This chapter discusses the development of research problems. We begin by clarifying some relevant terms. Basic Terminology At a general level, a researcher selects a topic or a phenomenon on which to focus. Examples of research topics are claustrophobia during MRI tests, pain management for sickle cell disease, and nutrition during pregnancy. Within broad topic areas are many potential research problems. In this section, we illustrate various terms using the topic side ef ects of chemotherapy. A research problem is an enigmatic or troubling condition. Researchers identify a research problem within a broad topic of interest. The purpose of research is to “solve” the problem—or to contribute to its
solution—by generating relevant, high-quality evidence. Researchers articulate the problem in a problem statement that also presents a rationale for the study. Many reports include a statement of purpose (or purpose statement), which summarizes the goal of the study. Research questions are the specific queries researchers want to answer in addressing the problem. Research questions guide the types of data to collect in a study. Researchers who make predictions about answers to research questions pose hypotheses that are tested in the study. These terms are not always consistently defined in research methods textbooks, and differences among them are often subtle. Table 4.1 illustrates the terms as we define them.
TABLE 4.1 Example of Terms Relating to Research Problems
Term Example
Topic/focus Side effects of chemotherapy Research problem (simple problem statement) Nausea and vomiting are common side effects among patients on chemotherapy, and interventions to date have been only moderately successful in reducing these effects. One issue concerns the efficacy of alternative means of administering antiemetic therapies. Statement of purpose The purpose of the study is to test an intervention to reduce chemotherapy-induced side effects—specifically, to compare the effectiveness of patient-controlled and nurse-administered antiemetic therapy for controlling nausea and vomiting in patients on chemotherapy. Research questions What is the relative effectiveness of patient-controlled antiemetic therapy versus nurse-controlled antiemetic therapy with regard to (1) medication consumption and (2) control of nausea and vomiting in patients on chemotherapy? Hypotheses Patients receiving antiemetic therapy by a patient-controlled pump will (1) be less nauseous, (2) vomit less, and (3) consume less medication than patients receiving the therapy by nurse administration.
Box 4.1 Draft Problem Statement on Humor and Stress A diagnosis of cancer is associated with high levels of stress. Sizable numbers of patients who receive a cancer diagnosis describe feelings of uncertainty, fear, anger, and loss of control. Interpersonal relationships, psychological functioning, and role performance have all been found to suffer following cancer diagnosis and treatment. A variety of alternative/complementary therapies have been developed in an effort to decrease the harmful effects of stress on psychological and physiological functioning, and resources devoted to these
therapies (money and staff) have increased in recent years. However, many of these therapies have not been carefully evaluated to determine their efficacy, safety, or cost-effectiveness. For example, the use of humor has been recommended as a therapeutic device to improve quality of life, decrease stress, and perhaps improve immune functioning, but the evidence to support this claim is limited.
Box 4.2 Some Possible Improvements to Problem Statement on Humor and Stress
Each year, more than 1 million people are diagnosed with cancer, which remains one of the top causes of death among both men and women (reference citations). Numerous studies have documented that a diagnosis of cancer is associated with high levels of stress. Sizable numbers of patients who receive a cancer diagnosis describe feelings of uncertainty, fear, anger, and loss of control (citations). Interpersonal
relationships, psychological functioning, and role performance have all been found to suffer following cancer diagnosis and treatment (citations). These stressful outcomes can, in turn, adversely affect health,
long-term prognosis, and medical costs among cancer survivors (citations). A variety of alternative/complementary therapies have been developed in an effort to decrease the harmful effects of stress on psychological and physiological functioning, and resources devoted to these
therapies (money and staff) have increased in recent years (citations). However, many of these therapies have not been carefully evaluated to determine their efficacy, safety, or cost-effectiveness. For example, the use of humor has been recommended as a therapeutic device to improve quality of life, decrease stress, and perhaps improve immune functioning (citations), but the evidence to support this claim is limited. Preliminary findings from a recent small-scale endocrinology study with a healthy sample exposed to a humorous intervention (citation) holds promise for further inquiry with immuno-compromised
Box 4.3 Guidelines for Critically Appraising Research Problems, Research Questions, and Hypotheses
1. What is the research problem? Is the problem statement easy to locate and is it clearly stated? Does the problem statement build a cogent and persuasive argument for the new study? 2. Does the problem have significance for nursing? How might the research contribute to nursing practice, administration, education, or policy? 3. Is there a good fit between the research problem and the paradigm in which the research was conducted? Is there a good fit between the problem and the qualitative research tradition (if applicable)? 4. Does the report formally present a statement of purpose, research question, and/or hypotheses? Is this information communicated clearly and concisely, and is it placed in a logical and useful location? 5. Are purpose statements or questions worded appropriately? For example, are key concepts/variables identified and is the population of interest specified? Are verbs used appropriately to suggest the nature of the inquiry and/or the research tradition? 6. If there are no formal hypotheses, is their absence justified? Are statistical tests used in analyzing the data despite the absence of stated hypotheses? 7. Do hypotheses (if any) flow from a theory or previous research? Is there a justifiable basis for the predictions? 8. Are hypotheses (if any) properly worded—do they state a predicted relationship between two or more variables? Are they directional or nondirectional, and is there a rationale for how they were stated? Are they presented as research or as null hypotheses?
Research Problems and Paradigms
Some research problems are better suited to qualitative versus quantitative methods. Quantitative studies usually focus on concepts that are fairly well developed, about which there is existing evidence, and for which reliable methods of measurement have been (or can be) developed. For example, a quantitative study might be undertaken to explore whether older people with chronic illness who continue working are
less (or more) depressed than those who retire. There are relatively good measures of depression that would yield quantitative information about the level of depression in a sample of employed and retired
seniors who are chronically ill. Qualitative studies are often undertaken because a researcher wants to develop a rich and context-bound understanding of a poorly understood phenomenon. Researchers often initiate a qualitative study to heighten awareness and create a dialogue about a phenomenon. Qualitative methods would not be well suited to comparing levels of depression among employed and retired seniors, but they would be ideal for exploring, for example, the meaning or experience of depression among chronically ill retirees. Thus, the nature of the research question is linked to paradigms and to research traditions within paradigms.
Sources of Research Problems Where do ideas for research problems come from? At a basic level, research topics originate with researchers’ interests. Because research is a time-consuming enterprise, curiosity about and interest in a topic are
essential. Research reports rarely indicate the source of researchers’ inspiration, but a variety of explicit sources can fuel their interest, including the following:
Clinical experience. Nurses’ everyday clinical experience is a rich source of ideas for research inquiries. Immediate problems that need a solution—analogous to problem-focused triggers discussed in Chapter 2—may generate enthusiasm and have high potential for clinical
relevance. Patients’ involvement. Increasingly, researchers are turning to patients and other key stakeholders for input in identifying important issues for research. Patient-centered outcomes research (PCOR) has become increasingly prominent. Quality improvement ef orts. Important clinical questions sometimes emerge in the context of findings from quality improvement studies. Personal involvement on a quality improvement team can sometimes lead to ideas for a study. In Chapter 12, we discuss a process called
root cause analysis that can suggest a research focus. Nursing literature. Ideas for studies sometimes come from reading the nursing literature. Research articles may suggest problems indirectly by stimulating the reader’s curiosity and directly by pointing out needed research. Social issues. Topics are sometimes suggested by global social or political issues of relevance to the healthcare community. For example, the feminist movement raised questions about such topics as gender equity in health care. Public awareness about health disparities has led
to research on healthcare access and culturally sensitive interventions.
Ideas from external sources. External sources and direct suggestions can sometimes provide the impetus for a research idea. For example, ideas for studies may emerge from brainstorming with other nurses. Additionally, researchers who have developed a program of research on a topic area may get inspiration for “next steps” from their own findings or from a discussion of those findings with others.
Example of a Problem Source in a Program of Research
Beck, one of this book’s authors, conducted a study with two collaborators (Beck et al., 2015) on secondary traumatic stress among certified nurse midwives (CNMs). Beck has developed a strong research program on postpartum depression and traumatic births. She and Gable had previously conducted a study with labor and delivery nurses and their experiences of secondary traumatic stress caring for women during traumatic births. When Beck presented the findings of this study at conferences, certified CNMs in the audience often said “You should research us too. We also have secondary traumatic
TIP Personal experiences in clinical settings are a provocative source of research ideas and questions. Here are some hints: Watch for a recurring problem and see if you can discern a pattern in situations that lead to the problem. Example: Why do so many patients complain of being tired after being transferred from a coronary care unit to a progressive care unit?
Think about aspects of your work that are frustrating or do not result in the intended outcome—then try to identify factors contributing to the problem that could be changed. Example: Why is suppertime so frustrating in a nursing home?
Critically examine your own clinical decisions. Are they based on tradition, or are they based on systematic evidence that supports their efficacy? Example: What would happen if you used the return of flatus to assess the return of GI motility after abdominal surgery, rather than listening to bowel sounds?
Developing and Refining Research Problems
Procedures for developing a research problem are difficult to describe. The process is rarely a smooth and orderly one; there are likely to be false starts, inspirations, and setbacks. The few suggestions offered here are not intended to imply that there are techniques for making this first step easy but rather to encourage you to persevere in the absence of instant success.
Selecting a Topic Developing a research problem is a creative process—and it is a process that is sometimes best done in teams. The teams can include other nurses, mentors, interdisciplinary partners, patients, or other community members.
In the early stages of initiating research ideas, try not to be too self-critical. It is better to relax and jot down topics of interest as they come to mind. It does not matter if the ideas are abstract or concrete, broad or specific, technical or colloquial—the important point is to put ideas on paper. After this first step, ideas can be sorted in terms of interest, knowledge about the topics, and the perceived feasibility of turning the ideas into a study. When the most fruitful topic area has been selected, the list should not be discarded; it may be necessary to return to it.
TIP The process of selecting and refining a research problem usually takes longer than you might think. The process involves starting with some preliminary ideas; having discussions with colleagues, advisers, or stakeholders; perusing the research literature; looking at what is happening in clinical settings; and a lot of reflection.
Narrowing the Topic Once you have identified a topic of interest, you can begin to ask some broad questions that can lead you to a researchable problem. Examples of question stems that might help to focus an inquiry include the
following: What is going on with …? What is the process by which …? What is the meaning of …? What would happen if …? What influences or causes …? What are the consequences of…? What factors contribute to …?
Early criticism of ideas can be counterproductive. Try not to jump to the conclusion that an idea sounds trivial or uninspired without giving it more careful consideration or exploring it with others. Another potential danger is that new researchers sometimes develop problems that are too broad in scope. The transformation of a general topic into a workable problem often is accomplished in uneven steps. Each step
should result in progress toward the goals of narrowing the scope of the problem and sharpening the concepts. As researchers move from general topics to more specific ideas, several possible research problems may emerge. Consider the following example. Suppose you were working on a medical unit and were puzzled
by that fact that some patients always complained about having to wait for pain medication when certain nurses were assigned to them. The general problem is discrepancy in patient complaints regarding pain medications. You might ask: What accounts for the discrepancy? How can I improve the situation? These are not research questions, but they may lead you to ask such questions as the following: How do the two groups of nurses differ? or What characteristics do the complaining patients share? At this point, you may observe that the cultural and ethnic background of the patients and nurses could be relevant. This may
lead you to search the literature for studies about culture and ethnicity in relation to nursing care, or it may prompt you to discuss your observations with others. These efforts may result in several research
questions, such as the following: What is the nature of patient complaints among patients of different cultural backgrounds?
Is the cultural background of nurses related to the frequency with which they dispense pain medication? Does the number of patient complaints increase when patients are of dissimilar cultural backgrounds as opposed to when they are of the same cultural background as nurses? Do nurses’ dispensing behaviors change as a function of the similarity between their own cultural background and that of patients?
These questions stem from the same problem, yet each would be studied differently. Some suggest a qualitative approach and others suggest a quantitative one. A quantitative researcher might be curious about cultural or ethnic differences in nurses’ dispensing behaviors. Both ethnicity and nurses’ dispensing behaviors are variables that can be operationalized. A qualitative researcher would likely be more interested in understanding the essence of patients’ complaints, their experience of frustration, or the process by which the problem got resolved. Researchers choose a problem to study based on several factors, including its inherent interest and its compatibility with a paradigm of preference. In addition, tentative problems vary in their feasibility and worth. A critical evaluation of ideas is appropriate at this point.
Evaluating Research Problems Although there are no rules for selecting a research problem, four important considerations to keep in mind are the problem’s significance, researchability, feasibility, and interest to you.
Significance of the Problem
A crucial factor in selecting a problem is its significance to nursing. Evidence from the study should have potential to contribute meaningfully to nursing; the new study should be the right “next step” in building
an evidence base. The right next step could be an original study, but it could also be a replication to answer previously asked questions with greater rigor or with a different population.
TIP In evaluating the significance of an idea, ask the following kinds of questions: Is the problem important to nursing and its clients? Will patient care benefit from the evidence? Will the findings challenge
(or lend support to) existing practices? If the answer to all these questions is “no,” then the problem probably should be abandoned.
Researchability of the Problem
Not all problems are amenable to research inquiry. Questions of a moral or ethical nature, although provocative, cannot be researched. For example, should assisted suicide be legalized? There are no right or wrong answers to this question, only points of view. Of course, related questions could be researched, such as: Do patients living with high levels of pain hold more favorable attitudes toward assisted suicide than
those with less pain? What moral dilemmas are perceived by nurses who might be involved in assisted suicide? The findings from studies addressing such questions would have no bearing on whether assisted
suicide should be legalized, but they could be useful in developing a better understanding of key issues.
Feasibility of the Problem
A third consideration concerns feasibility, which encompasses several issues. Not all of the following factors are universally relevant, but they should be kept in mind in making a decision. Time. Most studies have deadlines or completion goals, so the problem must be one that can be studied in the allotted time. It is prudent to be conservative in estimating time for the various tasks because research
activities typically require more time than anticipated. Researcher experience. Ideally, the problem should relate to a topic about which you have some prior knowledge or experience. Also, beginning researchers should avoid problems that might require the development of a new measuring instrument or that demand complex analyses. Availability of study participants. In any study involving humans, researchers need to consider whether people with the desired characteristics will be available and willing to cooperate. Researchers may need
to put considerable effort into recruiting participants or may need to offer a monetary incentive. Cooperation of others. It may be necessary to gain entrée into an appropriate community or setting and to develop the trust of gatekeepers. In institutional settings (e.g., hospitals), access to clients, personnel, or
records requires authorization. Ethical considerations. A research problem may be unfeasible if the study would pose unfair or unethical demands on participants. The ethical issues discussed in Chapter 7 should be reviewed when considering
a study’s feasibility. Facilities and equipment. All studies have resource requirements, although needs are sometimes modest. It is prudent to consider what facilities and equipment will be needed and whether they will be available. Money. Monetary needs for studies vary widely, ranging from $100 or less for small student projects to hundreds of thousands of dollars for large-scale research. If you are on a limited budget, you should think
carefully about projected expenses before selecting a problem. Major categories of research-related expenditures include:
Personnel costs—payments to research assistants (e.g., for interviewing, coding, data entry, transcribing, statistical consulting) Participant costs—payments to participants as an incentive for their cooperation or to offset their expenses (e.g., parking, babysitting costs)
Supplies—paper, memory sticks, postage, and so forth Printing and duplication—costs for reproducing forms, questionnaires, and so on Equipment—computers and software, audio- or video-recorders, calculators, and the like Laboratory fees for the analysis of biophysiologic data Transportation costs (e.g., travel to participants’ homes)
TIP If your study involves testing a new procedure or intervention, you should also consider the feasibility of ultimately implementing it in real-world settings, should it prove effective. If the innovation
requires a lot of resources, there may be little interest in adopting it, even if it results in improvements.
Researcher Interest Even if a tentative problem is researchable, significant, and feasible, there is one more criterion: your own interest in the problem. Genuine curiosity about a research problem is an important prerequisite to a
successful study. A lot of time and energy are expended in a study; there is little sense devoting these resources to a project about which you are not enthusiastic.
TIP New researchers often seek suggestions about a topic area, and such assistance may be helpful in getting started. Nevertheless, it is unwise to be talked into a topic in which you have limited interest. If you do not find a problem appealing at the beginning of a study, you are likely to regret your choice later.
Communicating Research Problems and Questions
Every study needs a problem statement—an articulation of what is problematic and is the impetus for the research. Most research reports also present a statement of purpose, research questions, or hypotheses. Many people do not understand problem statements and may have trouble identifying them in a research article—not to mention developing one. A problem statement often begins with the very first sentence after the abstract. Specific research questions, purposes, or hypotheses appear later in the introduction. Typically, however, researchers begin their inquiry by identifying their research question and then develop
an argument in their problem statement to present the rationale for the new research. This section follows that sequence by describing statements of purpose and research questions, followed by a discussion of problem statements.
Statements of Purpose Many researchers articulate their research goals in a statement of purpose, worded declaratively. It is usually easy to identify a purpose statement because the word purpose is explicitly stated “The purpose of this
study was…”—although sometimes the words aim, goal, or objective are used instead, as in “The goal of this study was….”
In a quantitative study, a statement of purpose identifies the key study variables and their possible interrelationships, as well as the population of interest (i.e., the PICO elements).
Example of a Statement of Purpose From a Quantitative Study
“Aim: This study examined the effects of a music intervention on anxiety, depression, and psychosomatic symptoms of oncology nurses” (Ploukou & Panagopoulou, 2018, p. 77).
In this purpose statement for a Therapy question, the population (P) is oncology nurses. The aim is to assess whether a music intervention (I) compared with no music intervention (C)—which together comprise
the independent variable—has an effect on the nurses’ anxiety, depression, and psychosomatic symptoms, which are the dependent variables (the Os).
In qualitative studies, the statement of purpose indicates the key concept or phenomenon, and the people under study.
Example of a Statement of Purpose From a Qualitative Study The aims of this study were “to explore the experiences of adherence to endocrine therapy in women with breast cancer and their perceptions of the challenges they face in adhering to their medication”
(Iacorossi et al., 2018, p. E57).
This statement indicates that the central phenomenon in this study was the experiences of medication adherence and related challenges among women with breast cancer (P). The statement of purpose communicates more than just the nature of the problem. Researchers’ selection of verbs in a purpose statement suggests how they sought to solve the problem, or the state of knowledge on the topic. A study whose purpose is to explore or describe a phenomenon is likely an investigation of a little-researched topic, sometimes involving a qualitative approach. A purpose statement for a qualitative
study may also use verbs such as understand, discover, or develop. Statements of purpose in qualitative studies may “encode” the tradition of inquiry, not only through the researcher’s choice of verbs but also
through the use of “buzz words” associated with those traditions, as follows:
Grounded theory: Processes; social structures; social interactions Phenomenologic studies: Experience; lived experience; meaning; essence Ethnographic studies: Culture; roles; lifeways; cultural behavior Quantitative researchers also suggest the nature of the inquiry through their selection of verbs. A statement indicating that the study’s purpose is to test or evaluate something (e.g., an intervention) suggests an
experimental design. A study whose purpose is to examine or explore the relationship between two variables likely involves a nonexperimental design. Sometimes the verb is ambiguous: a purpose statement
indicating that an intent to compare could be referring to a comparison of alternative treatments (using an experimental approach) or a comparison of preexisting groups (using a nonexperimental approach). In
any event, verbs such as test, evaluate, and compare suggest an existing knowledge base and quantifiable variables. The verbs in a purpose statement should connote objectivity. A statement of purpose indicating that the study goal was to prove, demonstrate, or show something suggests a bias. The word determine should usually
be avoided as well because research methods almost never provide definitive answers to research questions.
TIP Unfortunately, some reports fail to state the study purpose clearly, leaving readers to infer the purpose from such sources as the title of the report. In other reports, the purpose may be difficult to find. Researchers often state their purpose toward the end of the report’s introduction.
Research Questions Research questions are sometimes direct rewordings of purpose statements, phrased interrogatively rather than declaratively, as in the following example:
Purpose: The purpose of this study was to assess the relationship between the functional dependence level of renal transplant recipients and their rate of recovery. Question: What is the relationship between the functional dependence level (I and C: higher versus lower levels) of renal transplant recipients (P) and their rate of recovery (O)? Questions have the advantage of simplicity and directness—they invite an answer and help to focus attention on the kinds of data needed to provide that answer. Some research reports thus omit a statement of purpose and state only research questions. Other researchers use a set of research questions to clarify or lend greater specificity to a global purpose statement.
Research Questions in Quantitative Studies
In Chapter 2, we discussed the framing of clinical foreground questions to guide an EBP inquiry. Many of the EBP question templates in Table 2.1 could yield questions to guide a study as well, but researchers
tend to conceptualize their questions in terms of their variables. Take, for example, the Therapy question in Table 2.1, which states, “In (Population), what is the effect of (Intervention) on (Outcome)?” A researcher would likely think of the question in these terms: “In (population), what is the effect of (independent variable) on (dependent variable)?” Thus, in quantitative studies research questions identify the population
(P) under study, the key study variables (I, C, and O components), and possible relationships among the variables. The variables are all quantifiable concepts. Most research questions concern relationships, so many quantitative research questions could be articulated using a general template: “In (population), what is the relationship between (independent variable or
IV) and (dependent variable or DV)?” Variations include the following:
Therapy/intervention: In (population), what is the effect of (IV: intervention versus an alternative) on (DV)? Prognosis: In (population), does (IV: presence of disease or illness versus its absence) affect or increase the risk of (DV: adverse consequences)? Etiology/harm: In (population), does (IV: exposure versus nonexposure) cause or increase the risk of (DV: disease, health problem)?
Clinical foreground questions for an EBP-focused search and a question for a study sometimes differ. As shown in Table 2.1, sometimes clinicians ask PICO questions about explicit comparisons (e.g., they want to
compare intervention A with intervention B) and sometimes they do not (e.g., they want to learn the effects of intervention A, compared with those of any other intervention or to the absence of an intervention, PIO questions). In a research question, there must always be a designated comparison because the independent variable must be operationally defined; this definition would articulate the specific “I” and “C” being studied.
TIP Research questions are sometimes more complex than clinical foreground questions for EBP. They may include, in addition to the independent and dependent variable, elements called moderator variables or mediating variables. A moderator variable is a variable that influences the strength or direction of a a relationship between two variables (e.g., a person’s age might moderate the effect of exercise on physical function). A mediating variable is one that acts like a “go-between” in a link between two variables (e.g., a smoking cessation intervention may affect smoking behavior through the
intervention’s effect on motivation to quit). The Supplement for this chapter on this book’s website describes the role of moderating and mediating variables in complex research questions.
Some research questions are primarily descriptive. As examples, here are some descriptive questions that could be addressed in a study on nurses’ use of humor: What is the frequency with which nurses use humor as a complementary therapy with hospitalized patients with cancer? What are the reactions of hospitalized cancer patients to nurses’ use of humor? What are the characteristics of nurses who use humor as a complementary therapy with hospitalized patients with cancer?
Is my Use of Humor Scale a reliable and valid measure of nurses’ use of humor with patients in clinical settings? Answers to such questions might, if addressed in a methodologically sound study, be useful in developing interventions for reducing stress in patients with cancer.
Example of a Research Question From a Quantitative Study Lechner and colleagues (2018) studied skin condition and skin care in German care facilities. Here is one research question: What is the prevalence of dry skin in nursing home residents and hospital patients and is
the prevalence higher in nursing homes or hospitals?
TIP The Toolkit section of Chapter 4 of the accompanying Resource Manual includes question templates in a Word document that can be “filled in” to generate many types of research questions for both
qualitative and quantitative studies.
Research Questions in Qualitative Studies Research questions for qualitative studies state the phenomenon of interest and the group or population of interest. Researchers in the various qualitative traditions vary in their conceptualization of what types of questions are important. Grounded theory researchers are likely to ask process questions, phenomenologists tend to ask meaning questions, and ethnographers generally ask descriptive questions about cultures. Special terms associated with the various traditions, noted previously, are likely to be incorporated into the research questions.
Example of a Research Question From a Phenomenologic Study What is the lived experience of children with spina bifida in the West Bank, Palestine (Nahal et al., 2019)? Not all qualitative studies are rooted in a specific research tradition. Many researchers use qualitative methods to describe or explore phenomena without focusing on cultures, meaning, or social processes.
Example of a Research Question From a Descriptive Qualitative Study
In their descriptive qualitative study, Dial and Holmes (2018) asked, “What are the successful self-care hygienic strategies that patients of size use to care for themselves at home?”
In qualitative studies, research questions may evolve over the course of the study. Researchers begin with a focus that defines the broad boundaries of the study, but the boundaries are not cast in stone. The boundaries “can be altered and, in the typical naturalistic inquiry, will be” (Lincoln & Guba, 1985, p. 228). The naturalist begins with a research question that provides a general starting point but does not prohibit discovery. The emergent nature of qualitative inquiry means that research questions can be modified as new data make it relevant to do so.
Problem Statements
Problem statements express the dilemma or troubling situation that needs investigation and that provide a rationale for a new inquiry. A good problem statement is a well-structured formulation of what is problematic, what “needs fixing,” or what is poorly understood. Problem statements, especially for quantitative studies, often have most of the following six components:
1. Problem identification: What is wrong with the current situation? 2. Background: What is the context of the problem that readers need to understand? 3. Scope of the problem: How big a problem is it? How many people are affected? 4. Consequences of the problem: What are the costs of not fixing the problem? 5. Knowledge gaps: What information about the problem is lacking? 6. Proposed solution: How would the proposed study contribute to the solution of the problem?
These components, taken together, form the argument for the study—researchers try to persuade readers that the rationale for undertaking the study is sound.
TIP The Toolkit section of Chapter 4 of the accompanying Resource Manual includes these six questions in a Word document that can be “filled in” and reorganized as needed, as an aid to developing a problem statement.
Suppose our topic was humor as a complementary therapy for reducing stress in hospitalized patients with cancer. Our research question is, “What is the effect of nurses’ use of humor on stress and natural killer cell activity in hospitalized patients with cancer?” Box 4.1 presents a rough draft of a problem statement for such a study. This problem statement is a reasonable first draft. The draft has several, but not all, of the
six components. Box 4.2 illustrates how the problem statement could be strengthened by adding information about scope (component 3), long-term consequences (component 4), and possible solutions (component 6). This second
draft builds a more compelling argument for new research: millions of people are affected by cancer, and the disease has adverse consequences not only for those diagnosed and their families but also for society. The revised problem statement also suggests a basis for the new study by describing a solution on which the new study might build. As this example suggests, the problem statement is usually interwoven with supportive evidence from the research literature. In many research articles, it is difficult to disentangle the problem statement from the
literature review, unless there is a subsection specifically labeled “Literature Review.” Problem statements for a qualitative study similarly express the nature of the problem, its context, its scope, and information needed to address it, as in the following abridged example:
Example of a Problem Statement From a Qualitative Study
“Rheumatoid arthritis (RA) and psoriatic arthritis (PsA) are inflammatory diseases characterised by chronic arthritis that can result in considerable disease burden. Disease activity and symptoms of RA and
PsA can contribute to reduced physical, emotional or psychosocial health and well-being…A physically active lifestyle is associated with reduced risk of several diseases…However, only a minority of people with RA participate in health-promoting physical activities…In addition, people with RA report high levels of pain-catastrophising exhibited as high levels of self-rated pain associated with increased
fear-avoidance behaviour towards physical activity…This study was conducted to gain better insight into fear-avoidance beliefs in relation to physical activity among people experiencing moderate-to-severe
rheumatic pain.” (Lööf & Johansson, 2019, p. 322). Qualitative studies embedded in a particular research tradition usually incorporate terms in their problem statements that foreshadow the tradition. For example, the problem statement in a grounded theory
study might refer to the need to generate a theory relating to social processes. A problem statement for a phenomenologic study might note the need to gain insight into people’s experiences or the meanings they
attribute to those experiences. And an ethnographer might indicate the need to understand how cultural forces affect people’s health behaviors.
Research Hypotheses
A hypothesis is a prediction, almost always a prediction about the relationship between variables. 1 In qualitative studies, researchers do not have an a priori hypothesis, in part because there is too little known to
justify a prediction and in part because qualitative researchers want the inquiry to be guided by participants’ viewpoints rather than by their own hunches. Thus, our discussion here focuses on hypotheses in
quantitative research.
Function of Hypotheses in Quantitative Research
Research questions, as we have seen, are usually queries about relationships between variables. Hypotheses are predicted answers to these queries. For instance, the research question might ask: Does sexual abuse in childhood affect the development of irritable bowel syndrome in women? The researcher might predict the following: Women (P) who were sexually abused in childhood (I) have a higher incidence of
irritable bowel syndrome (O) than women who were not (C). Hypotheses sometimes follow from a theory. Scientists reason from theories to hypotheses and test those hypotheses in the real world. Take, as an example, the theory of reinforcement, which maintains that behavior that is positively reinforced (rewarded) tends to be learned or repeated. Predictions based on this theory could be tested. For example, we could test the following hypothesis: Pediatric patients (P) who
are given a reward (e.g., a toy) (I) when they undergo nursing procedures tend to be more cooperative during those procedures (O) than nonrewarded peers (C). This hypothesis can be put to a test, and the
theory gains credibility if it is supported with real data. Even in the absence of a theory, well-conceived hypotheses offer direction and suggest explanations. For example, suppose we hypothesized that cue-based feedings compared with traditional methods of feeding
for preterm infants will shorten the time to full oral feedings and discharge from the NICU. We could justify our speculation based on earlier studies or clinical observations, or both. The development of predictions
forces researchers to think logically and to tie together earlier research findings. Now let us suppose the preceding hypothesis is not confirmed: we find that time to full oral feedings and discharge is similar for preterm infants on cue-based feedings and traditional methods of feeding. The
failure of data to support a prediction forces researchers to analyze theory or previous research critically, to consider study limitations, and to explore alternative explanations for the findings. The use of hypotheses tends to induce
critical thinking and encourages careful interpretation of the evidence. To illustrate further the utility of hypotheses, suppose we conducted the study guided only by the research question, Is there a relationship between feeding method in preterm infants and the length of time to
full oral feedings and NICU discharge? The investigator without a hypothesis is apparently prepared to accept any results. The problem is that it is almost always possible to explain something superficially after
the fact, no matter what the findings are. Hypotheses reduce the risk that spurious results will be misconstrued.
TIP Consider whether it might be appropriate to develop hypotheses that predict different effects of the independent variable on the outcome for different subgroups of people—that is, to consider the
effects of moderator variables. For example, would you predict the effects of an intervention to be different for males and females? Testing such hypotheses might facilitate greater applicability of the evidence
to specific types of patient (Chapter 31).
Characteristics of Testable Hypotheses
Testable hypotheses state the expected relationship between the independent variable (the presumed cause or antecedent) and the dependent variable (the presumed effect or outcome) within a population.
Example of a Research Hypothesis Palesh and colleagues (2018) hypothesized that, among women with advanced breast cancer, a greater degree of physical activity is associated with longer survival.
In this example, the population is women with advanced breast cancer, the independent variable is amount of physical activity, and the dependent variable is length of time before death. The hypothesis predicts
that these two variables are related within the population—greater physical activity is predicted to be associated with longer survival. Hypotheses that do not make a relational statement are difficult to test. Take the following example: Pregnant women who receive prenatal instruction about postpartum experiences are not likely to experience postpartum
depression. This statement expresses no anticipated relationship—there is only one variable (postpartum depression), and a relationship requires at least two variables. The problem is that without a prediction about an anticipated relationship, the hypothesis is difficult to test using standard statistical procedures. In our example, how would we know whether the hypothesis was supported—what standard could be used to decide whether to accept or reject it? To illustrate this concretely, suppose we asked a group of mothers who had been given instruction on postpartum
experiences the following question 1 month after delivery: On the whole, how depressed have you been since you gave birth? Would you say (1) extremely depressed, (2) moderately depressed, (3) a little depressed, or (4) not at all depressed? Based on responses to this question, how could we compare the actual outcome with the predicted outcome? Would all the women have to say they were “not at all depressed?” Would the prediction be
supported if 51% of the women said they were “not at all depressed” or “a little depressed?” It is difficult to test the accuracy of the prediction. A test is simple, however, if we modify the prediction as follows: Pregnant women who receive prenatal instruction are less likely to experience postpartum depression than those with no prenatal instruction. Here, the outcome variable (O) is the women’s depression, and the independent variable is receipt (I) versus nonreceipt (C) of prenatal instruction. The relational aspect of the prediction is embodied in the phrase
less than. If a hypothesis lacks a phrase such as more than, less than, greater than, dif erent from, related to, associated with, or something similar, it is probably not amenable to statistical testing. To test this revised
hypothesis, we could ask two groups of women with different prenatal instruction experiences to respond to the question on depression and then compare the average responses of the two groups. The absolute degree of depression of either group would not be at issue. Hypotheses should be based on justifiable rationales. Hypotheses often follow from previous research findings or are deduced from a theory. When a new area is being investigated, the researcher may have to
turn to logical reasoning or clinical experience to justify predictions.
The Derivation of Hypotheses Many students ask, How do I go about developing hypotheses? Two basic processes—induction and deduction—are the intellectual machinery involved in deriving hypotheses (The Supplement to Chapter 3 on the Point book’s website described induction and deduction). An inductive hypothesis is inferred from observations. Researchers observe certain patterns among phenomena and then make predictions based on the observations. An important source for inductive hypotheses is clinical experiences. For example, a nurse might notice that presurgical patients who ask a lot of questions about pain have a more difficult time than other patients in learning postoperative
procedures. The nurse could formulate a hypothesis, such as: Patients who are stressed by fear of pain have more difficulty in deep breathing and coughing after surgery than patients who are not stressed. Qualitative studies are an important source of inspiration for inductive hypotheses.
Example of Deriving an Inductive Hypothesis LoGiudice and Beck (2016) conducted a phenomenological study of the experience of childbearing from eight survivors of sexual abuse. One of the themes from this study was “Overprotection: Keeping my
child safe.” A hypothesis that can be derived from this qualitative finding might be as follows: Women who are survivors of sexual abuse will be more overprotective of their children than mothers who have not experienced sexual abuse.
Inductive hypotheses begin with specific observations and move toward generalizations. Deductive hypotheses have theories or prior knowledge as a starting point, as in our earlier example about reinforcement
theory. Researchers deduce that if the theory is true, then certain outcomes can be expected. If hypotheses are supported, then the theory is strengthened. The advancement of nursing knowledge depends on both
inductive and deductive hypotheses. Researchers need to be organizers of concepts (think inductively), logicians (think deductively), and critics and skeptics of resulting formulations, constantly demanding
evidence. Wording of Hypotheses A good hypothesis is worded clearly and concisely and in the present tense. Researchers make predictions about relationships that exist in the population and not just about a relationship that will be revealed in a particular sample. There are various types of hypotheses.
Directional Versus Nondirectional Hypotheses Hypotheses can be stated in a number of ways, as in the following example:
1. Older patients are more likely to fall than younger patients. 2. There is a relationship between the age of a patient and the risk of falling. 3. The older the patient, the greater the risk that he or she will fall. 4. Older patients differ from younger ones with respect to their risk of falling. 5. Younger patients tend to be less at risk of a fall than older patients.
In each example, the hypothesis indicates the population (patients), the independent variable (patients’ age), the dependent variable (a fall), and the anticipated relationship between them. Hypotheses can be either directional or nondirectional. A directional hypothesis is one that specifies not only the existence but the expected direction of the relationship between variables. In our example, hypotheses 1, 3, and 5 are directional because there is an explicit prediction that older patients are more likely to fall than younger ones. A nondirectional hypothesis does not state the direction of the
relationship, as illustrated by versions 2 and 4. These hypotheses predict that a patient’s age and risk of falling are related, but they do not stipulate whether the researcher thinks that older patients or younger ones are at greater risk. Hypotheses derived from theory are almost always directional because theories provide a rationale for expecting variables to be related in a certain way. Existing studies also offer a basis for directional hypotheses. When there is no theory or related research, when findings of prior studies are contradictory, or when researchers’ own experience leads to ambivalence, nondirectional hypotheses may be appropriate. Some people argue, in fact, that nondirectional hypotheses are preferable because they connote impartiality. Directional hypotheses, it is said, imply that researchers are intellectually committed to
certain outcomes, and such a commitment might lead to bias. Yet, researchers typically do have hunches about outcomes, whether they state them explicitly or not. We prefer directional hypotheses when there is a reasonable basis for them because they clarify the study’s framework and demonstrate that researchers have thought critically about the study variables.
TIP Hypotheses can be either simple hypotheses (ones with one independent variable and one dependent variable) or complex hypotheses (ones with three or more variables—for example, with multiple
independent or dependent variables). Information about complex hypotheses is available in the Supplement for this chapter on .
Research Versus Null Hypotheses Hypotheses can be described as either research hypotheses or null hypotheses. Research hypotheses (also called scientific hypotheses) are statements of expected relationships between variables. All hypotheses presented thus far are research hypotheses that state actual predictions. Statistical inference uses a logic that may be confusing. This logic requires that hypotheses be expressed as an expected absence of a relationship. Null hypotheses (or statistical hypotheses) state that there is no
relationship between the independent and dependent variables. The null form of the hypothesis used in our example might be: “Patients’ age is unrelated to their risk of falling” or “Older patients are just as
likely as younger patients to fall.” The null hypothesis might be compared with the assumption of innocence of an accused criminal in many justice systems: the variables are assumed to be “innocent” of any
relationship until they can be shown “guilty” through appropriate statistical procedures. The null hypothesis represents the formal statement of this assumption of “innocence.” Researchers typically state research rather than null hypotheses. Indeed, you should avoid stating hypotheses in null form in a proposal or a report because this gives an amateurish impression. In statistical
testing, underlying null hypotheses are assumed without being stated. If the researcher’s actual research hypothesis is that no relationship among variables exists, complex procedures are needed to test it. Hypothesis Testing and Proof Hypotheses are formally tested through statistical analysis. Researchers use statistics to test whether their hypotheses have a high probability of being correct (i.e., have a p < .05). Statistical analysis does not offer proof; it only supports inferences that a hypothesis is probably correct (or not). Hypotheses are never proved or disproved; rather, they are supported or rejected. Findings are always tentative. Hypotheses come to be
increasingly supported with evidence from multiple studies. Let us look at why this is so. Suppose we hypothesized that height and weight are related. We predict that, on average, tall people weigh more than short people. We then obtain height and weight measurements
from a sample and analyze the data. Now suppose we happened by chance to get a sample that consisted of short, heavy people, and tall, thin people. Our results might indicate that there is no relationship
between height and weight. But we would not be justified in concluding that this study proved or demonstrated that height and weight are unrelated. This example illustrates the difficulty of using observations from a sample to draw definitive conclusions about a population. Issues such as the accuracy of the measures, the effects of uncontrolled variables, and
idiosyncracies of the study sample prevent researchers from concluding that hypotheses are proved.
TIP If a researcher uses any statistical tests (as is true in most quantitative studies), it means that there were underlying hypotheses—regardless of whether the researcher explicitly stated them—because
statistical tests are designed to test hypotheses. In planning a quantitative study of your own, do not hesitate to state hypotheses.
Critical Appraisal of Research Problems, Research Questions, and Hypotheses
In appraising research articles, you need to evaluate whether researchers have adequately communicated their problem. The problem statement, purpose, research questions, and hypotheses set the stage for a description of what the researchers did and what they learned. You should not have to dig deeply to decipher the research problem or the questions. A critical appraisal of the research problem is multidimensional. Substantively, you need to consider whether the problem has significance for nursing. Studies that build in a meaningful way on existing
knowledge are well-poised to contribute to evidence-based nursing practice. Researchers who develop a systematic program of research, designing new studies based on their own earlier findings, are especially
likely to make important contributions (Conn, 2004). For example, Cheryl Beck’s series of studies relating to postpartum depression and traumatic births have influenced women’s health care worldwide. Also,
research problems stemming from established research priorities (Chapter 1) have a high likelihood of yielding important new evidence for nurses because they reflect expert opinion about areas of needed
research. Another dimension in appraising the research problem is methodologic—in particular, whether the research problem is compatible with the chosen research paradigm and its associated methods. You should also
evaluate whether the statement of purpose or research questions have been properly worded and lend themselves to empirical inquiry.
In a quantitative study, if the research article does not contain explicit hypotheses, you need to consider whether their absence is justified. If there are hypotheses, you should evaluate whether they are logically
connected to the problem and are consistent with existing evidence or relevant theory. The wording of hypotheses should also be assessed. To be testable, the hypothesis should contain a prediction about the
relationship between two or more measurable variables. Specific guidelines for critically appraising research problems, research questions, and hypotheses are presented in Box 4.3.
Research Examples
This section describes how the research problem and research questions were communicated in two nursing studies, one quantitative and one qualitative.
Research Example of a Quantitative Study
Study: Effectiveness of a patient-centred, empowerment-based intervention programme among patients with poorly controlled type 2 diabetes (Cheng et al., 2018). Problem statement (Excerpt; citations omitted to streamline presentation): “Despite extensive advances and collective prioritization of evidence-based diabetes management, poor glycaemic control still remains common in many countries…Adherence to diabetes self-management regimen continues to be the most significant determinant to attain glycaemic target. Patients with poorly controlled type 2 diabetes find
enormous difficulty synthesizing self-management recommendations in the dynamic and complex daily context. There is a great call to support and empower them to take a proactive self-management role in the disease trajectory. A flourishing body of studies have illustrated that patient-centred, empowerment-based approach could boost patients’ engagement in and commitment to diabetes self-management.” (p. 44). Statement of purpose: The aim of this study was “to evaluate the effectiveness of a patient-centred, empowerment-based programme on glycaemic control and self-management behaviours among patients with poorly controlled type 2 diabetes.” (p.43). Research question: Although not formally stated by the researchers, we can state their Therapy question as follows: Among patients with poorly controlled type 2 diabetes (P), does participation in a patient– centered self-management program (I), compared with nonparticipation (C), lead to improvements in HbA1c levels and self-management behaviors (O)? Hypotheses: The researchers hypothesized that compared with study participants who do not receive the intervention, patients who receive the intervention program will have (1) significantly optimized
glycaemic control and (2) better self-management behaviors. Study methods: The study was conducted in two tertiary hospitals in China. A total of 242 eligible patients were recruited and were allocated, at random, to either receive or not receive the intervention. Those in
the intervention group received a 6-week self-management program; the control group received general health education and post discharge follow-up. The key outcomes were HbA1c levels and scores on a measure of self-management behaviors. Key findings: HbA1c values declined in both groups, and group differences at follow-up were not statistically significant. However, patients in the intervention group exhibited significant improvements in diet management and blood glucose self-monitoring both in the short term (8-week follow-up) and longer term (20-week follow-up).
Research Example of a Qualitative Study
Study: Patients’ perceptions and experiences of living with a surgical wound healing by secondary intention (McCaughan et al., 2018). Problem statement (Excerpt; citations omitted to streamline presentation): Most surgeries in the United Kingdom “result in a wound that heals by primary intention; that is to say, the incision is closed by fixing
the edges together with sutures (stitches), staples, adhesive glue, or clips. However, some wounds may be left open to heal…Healing occurs through the growth of new tissue from the base of the wound
upwards, a process described as ‘healing by secondary intention.’ …Management of open surgical wounds requires intensive treatment that may involve prolonged periods of hospitalisation for patients and/or
further surgical intervention…While there is an expansive literature relating to patients’ experiences of chronic wounds, such as leg ulcers, evidence concerning the impact on patients of experiencing an open
surgical wound is lacking.” (p. 30). Statement of purpose: The objective of this study was “to explore patients’ views and experiences of living with a surgical wound healing by secondary intention” (p. 29). Research questions: The patients’ experiences were explored by asking such questions as “How has this wound impacted on your daily life?” and “What effect has the wound had on your relationship with your
immediate family or friends?” Method: 20 patients from two locations in the north of England who had a surgical wound healing by secondary intention participated in the study. The researchers made efforts to recruit patients of different gender, age, wound duration, and type of surgery. The study was designed in collaboration with three patient advisers. Study participants were interviewed in-depth, with interviews continuing until data
saturation occurred. Key findings: The patients reported that alarm, shock, and disbelief were their initial reactions to their surgical wound healing. Wound-associated factors had a profound negative impact on their daily life, physical and psychosocial functioning, and well-being. Feelings of powerlessness and frustration were common, and many expressed dissatisfaction with the perceived lack of continuity of care in relation to wound management.
Summary Points
A research problem is a perplexing or enigmatic situation that a researcher wants to address through disciplined inquiry. Researchers usually identify a broad topic, narrow the problem scope, and identify questions consistent with a paradigm of choice. Common sources of ideas for nursing research problems are clinical experience, patient queries, relevant literature, quality improvement initiatives, social issues, and external suggestions. Key criteria in selecting a research problem are that the problem should be clinically important; researchable; feasible; and of personal interest. Feasibility involves the issues of time, researcher skills, cooperation of participants and other people, availability of facilities and equipment, adequacy of resources, and ethical considerations. Researchers communicate their goals as problem statements, statements of purpose, research questions, or hypotheses. Problem statements, which articulate the nature, context, and significance of a problem, include several components organized to form an argument for a new study: problem identification; the background, scope, and consequences of the problem; knowledge gaps; and
possible solutions to the problem. A statement of purpose, which summarizes the overall study goal, identifies key concepts or variables and the population. Purpose statements often communicate, through the use of verbs and other key terms, the underlying research tradition of qualitative studies, or whether study is experimental or nonexperimental in quantitative ones. A research question is the specific query researchers want to answer in addressing the research problem. In quantitative studies, research questions usually focus on relationships between variables.
In quantitative studies, a hypothesis is a statement of predicted relationships between two or more variables. Complex hypotheses may involve a moderator variable (a variable that alters the strength or direction of a relationship between two variables) or a mediating
variable that acts as a “go-between” in the link between two variables. Directional hypotheses predict the direction of a relationship; nondirectional hypotheses predict the existence of relationships, not their direction. Research hypotheses predict the existence of relationships; null hypotheses, which express the absence of a relationship, are the hypotheses subjected to statistical testing. Hypotheses are never proved or disproved in an ultimate sense—they are accepted or rejected, supported or not supported by the research data.
Study Activities
Study activities are available to instructors on .
References Cited in Chapter 4
Beck C. T., LoGiudice J., & Gable R. K. (2015). A mixed methods study of secondary traumatic stress in certified nurse-midwives: shaken belief in the birth process. Journal of Midwifery & Women’s Health, 60, 16–23. Cheng L., Sit J., Choi K., Chair S., Li X., Wu Y., … Tao M. (2018). Effectiveness of a patient-centred, empowerment-based intervention programme among patients with poorly controlled type 2 diabetes. International Journal of Nursing Studies, 79, 43–51. Conn V. (2004). Building a research trajectory. Western Journal of Nursing Research, 26, 592–594. Dial M., & Holmes J. (2018). “I do the best I can;” Personal care preferences of patients of size. Applied Nursing Research, 39, 259–264.
Iacorossi L., Gambalunga F., Fabi A., Giannarelli D., Marchetti A., Piredda M., & DeMarinis M. (2018). Adherence to oral administration of endocrine treatment in patients with breast cancer. Cancer Nursing, 41, E57–E63.
* Lechner A., Lahmann N., Lichterfeld-Kottner A., Müller-Werdan U., Blume-Peytavi U., & Kottner J. (2018). Dry skin and the use of leave-on products in nursing care: a prevalence study in nursing homes and hospitals. Nursing Open, 6, 189–196. Lincoln Y. S., & Guba E. G. (1985). Naturalistic inquiry. Newbury Park, CA: Sage. LoGiudice J. A., & Beck C. T. (2016). The lived experience of childbearing from survivors of sexual abuse: “it was the best of times, it was the worst of times”. Journal of Midwifery & Women’s Health, 61, 474–481. Lööf H., & Johansson U. (2019). “A body in transformation”—an empirical phenomenological study about fear-avoidance beliefs toward physical activity among persons experiencing moderate to severe rheumatic pain. Journal of Clinical Nursing, 28, 321–329. McCaughan D., Sheard L., Cullum N., Dunville J., & Chetter I. (2018). Patients’ perceptions and experiences of living with a surgical would healing by secondary intention. International Journal of Nursing Studies, 77, 29–38. Nahal M., Axelsson A., Iman A., & Wigert H. (2019). Palestinian children’s narratives about living with spina bifida: stigma, vulnerability, and social exclusion. Child: Care, Health, and Development, 45, 54–62.
** Palesh O., Kamen C., Sharp S., Golden A., Neri E., Spiegel D., & Koopman C. (2018). Physical activity and survival in women with advanced breast cancer. Cancer Nursing, 41, E31–E38. Ploukou S., & Panagopoulou E. (2018). Playing music improves well-being of oncology nurses. Applied Nursing Research, 39, 77–80.
*A link to this open-access article is provided in the Toolkit for Chapter 4 in the Resource Manual.
**This journal article is available on for this chapter.
1Although this does not occur with great frequency, it is possible to make a hypothesis about a specific value. For example, we might hypothesize that the rate of medication compliance in a specific population is 60%. Chapter 18 has an example.
Literature Reviews: Finding and Critically Appraising Evidence
A research literature review is a written synthesis and appraisal of evidence on a research problem. Researchers typically undertake a literature review as an early step in conducting a study. This chapter describes activities associated with literature reviews, including locating and critically appraising studies.
Some Literature Review Basics
Before discussing the steps involved in doing a research-based literature review, we briefly discuss some general issues.
Purposes of Research Literature Reviews Healthcare professionals are undertaking many different types of research synthesis, several of which are specifically intended to support evidence-based practice. Grant and Booth (2009) identified 14 different
types of synthesis—and even more review types are now appearing in the literature. We described one type of synthesis (systematic reviews) in Chapter 2, and several others will be discussed in Chapter 30. In
this chapter, we focus primarily on narrative literature reviews that researchers prepare during the conduct of a new study.
TIP A narrative literature review is one in which the findings from the studies under review are integrated using the judgments of the reviewers, rather than through statistical integration—as in a meta– analysis. Until meta-analytic techniques were developed, all reviews were narrative reviews. Once a research problem and research questions have been identified, a thorough literature review is essential. Literature reviews provide researchers with information to guide a high-quality study, such as
information about the following:
The scope and complexity of the identified research problem (for the argument); What other researchers have found in relation to the research question; The quality and quantity of existing evidence; The contexts and locales in which research has been conducted; The characteristics of the people who have served as study participants; Theoretical underpinnings of completed studies; Methodologic strategies that have been used to address the question; and Gaps in the existing evidence base—the type of new evidence that is needed.
This list suggests that a good literature review requires thorough familiarity with available evidence. As Garrard (2017) has advised, you must strive to own the literature on a topic to be confident of preparing a high-quality review. The term “reviewing the literature” is often used to refer to the process of identifying, locating, and reading relevant sources of research evidence—that is, conducting a literature review. However, researchers will ultimately need to summarize what they have learned in written form. The length of the product depends on its purpose. Written narrative literature reviews may take the following forms:
A review embedded in a research report. Literature reviews in the introduction to a research report provide readers with an overview of existing evidence and contribute to the argument for new research. These reviews are usually only two to three double-spaced pages, and so
only key studies can be cited. The emphasis is on summarizing and critiquing an overall body of evidence and demonstrating the need for a new study. A review in a research proposal. A literature review in a proposal (often, to request financial support) provides context and illuminates the rationale for new research. The length of such reviews is specified in proposal guidelines; sometimes it is just a few pages. When this is the
case, the review must reflect expertise on the topic in a succinct fashion. A review in a thesis or dissertation. Dissertations in the traditional format (see Chapter 32) often include a thorough, critical literature review. An entire chapter may be devoted to the review, and such chapters are often 20 to 30 pages long. These reviews typically include an
evaluation of the overall body of literature as well as critiques of key individual studies. They may also describe relevant theoretical foundations for the study.
In all three cases, the review is not simply a knowledge synthesis: the review provides a context for readers of the report or proposal and offers a justification for a new inquiry. Such reviews also can demonstrate
the researcher’s competence and thoroughness. Additionally, nurses sometimes prepare free-standing narrative reviews that are not necessarily done in connection with a planned new study. A written review may be undertaken as a course requirement in
graduate school or for publication in a journal. As an example, Gleason et al. (2018) published a literature review on the prevalence of atrial fibrillation symptoms and the relationship between such symptoms and
patients’ sex, race, and psychological distress. Such free-standing reviews are usually 15 to 25 pages long.
Literature Reviews in Qualitative Research Quantitative researchers almost always do an upfront literature review, but qualitative researchers have varying opinions about reviewing the literature before doing a new study. Some of the differences reflect viewpoints associated with qualitative research traditions. Grounded theory researchers often collect and analyze their data before reviewing the literature. Researchers turn to the literature once the grounded theory is sufficiently developed, seeking to relate the theory
to prior findings. Glaser (1978) warned that, “It’s hard enough to generate one’s own ideas without the ‘rich’ detailment provided by literature in the same field” (p. 31). Thus, grounded theory researchers may defer doing a literature review, but then later consider how previous research fits with or extends the emerging theory. Phenomenologists often undertake a search for relevant materials at the outset of a study, looking in particular for experiential descriptions of the phenomenon being studied (Munhall, 2012). The purpose is to
expand the researcher’s understanding of the phenomenon from multiple perspectives, and this may include an examination of artistic sources in which the phenomenon is described (e.g., in novels or poetry). Even though “ethnography starts with a conscious attitude of almost complete ignorance” (Spradley, 1979, p. 4), literature relating to the chosen cultural problem is often reviewed before data collection. A
second, more thorough literature review is often done during data analysis and interpretation so that findings can be compared with previous findings. Regardless of tradition, if funding is sought for a qualitative project, an upfront literature review is usually necessary. Proposal reviewers need to understand the context for a proposed study when deciding whether it should be funded.
Sources for a Research Review Written source materials vary in their quality and content. In performing a literature review, you will have to decide what to read and what to include in a written review. You may begin your search with broad
reference sources on a topic (e.g., textbooks), but ultimately you will mostly be retrieving information from articles published in professional journals. Findings from prior completed studies are the most important type of information for a research review. You should rely mostly on primary source research reports, which are descriptions of studies written by
the researchers who conducted them.
TIP Study protocols are an additional type of primary source—they are descriptions of the design and methods for studies that are underway but have not yet been completed. These protocols, which are available in registries and sometimes in journals, allow researchers to understand what new evidence will become available and hence can help you avoid unwanted duplication.
Secondary sources are descriptions of studies prepared by someone other than the original researcher. Literature reviews, for example, are secondary sources. If reviews are recent, they are very useful because
they provide an overview of the topic and a valuable bibliography. Secondary sources are not substitutes for primary sources because they typically fail to provide much detail about studies and may not be
completely objective.
In addition to research reports, your search may yield nonresearch references, such as case reports, anecdotes, editorials, or clinical descriptions. Nonresearch materials may broaden understanding of a problem, demonstrate a need for research, or describe aspects of clinical practice. These writings may help in formulating research ideas, but they usually have limited utility in written research reviews because they do not address the central question: What is the current state of evidence on this research problem?
Primary and Secondary Questions for a Review
For free-standing literature reviews, reviewers may summarize evidence about a single focused question, such as: Do virtual reality goggles (I) reduce pain (O) in patients undergoing wound care procedures (P)? For
those who are undertaking a literature review as part of a new study, the primary question for the literature review is the same as the research question for the new study. The researcher wants to know: What is the
current state of knowledge on the question that I will be addressing in my study?
If you are doing a review for a new study, you inevitably will need to search for current evidence on several secondary questions because you need to develop an argument for the new study. An example will clarify this point. Suppose that we were conducting a study to address the following question: Among nurses working in hospitals (P), what characteristics of the nurses or their practice settings (I) are associated with their management of children’s pain (O)? Such a question might arise in the context of a perceived problem, such as a concern that nurses’ treatment of children’s pain is not always optimal. A simplified statement of the problem might be as follows: Many children are hospitalized annually and many hospitalized children experience high levels of pain. Although effective analgesic and nonpharmacologic methods of controlling children’s pain exist, and
although there are reliable methods of assessing children’s pain, previous studies have found that nurses do not always manage children’s pain effectively. This rudimentary problem statement suggests a number of secondary questions for which up-to-date evidence needs to be found. Examples of such secondary questions include the following:
How many children are hospitalized each year? What levels of pain do hospitalized children typically experience? How can pain in hospitalized children be reliably assessed? How knowledgeable are nurses about pain assessment and pain management strategies for children?
Thus, a literature review tends to be a multipronged task when it is done in preparation for a new study. It is important to identify all questions for which information from the research literature needs to be
retrieved. Major Steps and Strategies in a Narrative Literature Review
Conducting a literature review is a little like doing a full study, in the sense that reviewers start with a question, formulate and implement a plan for gathering information, and then analyze and interpret the
information. The “findings” are then summarized in a written product. Figure 5.1 outlines key steps in the literature review process. As the figure shows, there are several potential feedback loops, with opportunities to retrace earlier steps in search of more information. This chapter
discusses each step, but some steps are elaborated in Chapter 30 in our discussion of systematic reviews.
FIGURE 5.1 Flow of tasks in a literature review.
Conducting a high-quality literature review is more than a mechanical exercise—it is an art and a science. Several qualities characterize a high-quality review. First, the review must be comprehensive, thorough, and up-to-date. To “own” the literature (Garrard, 2017), you must be determined to become an expert on your topic, which means that you need to be diligent in hunting down leads for possible sources of evidence.
TIP Locating all relevant information on a research question is like being a detective. The literature retrieval tools we discuss in this chapter are essential aids, but there inevitably needs to be some digging
for the clues to evidence on a topic. Be prepared for sleuthing.
Second, a high-quality review is systematic. Decision rules should be clear, and criteria for including or excluding a study need to be explicit. This is because a third characteristic of a good review is that it is
reproducible, which means that another diligent reviewer would be able to apply the same decision rules and criteria and come to similar conclusions about the evidence. Another desirable attribute of a literature review is the absence of bias. This is more easily achieved when systematic rules for evaluating information are followed or when a team of researchers participates in the
review—as is almost always the case in systematic reviews. Finally, reviewers should strive for a review that is insightful and that is more than “the sum of its parts.” Reviewers can contribute to knowledge
through an astute synthesis of the evidence. Doing a literature review is somewhat similar to doing a qualitative study: you will need a flexible and creative approach to “data collection.” Leads for relevant studies should be pursued until “saturation” is achieved—i.e., until your search strategies yield redundant information about studies to include. Finally, the analysis of your “data” will typically involve the identification of important themes in the literature. Organization in Literature Reviews
The importance of being well-organized in conducting a literature review cannot be overemphasized. As discussed in “Documentation in Literature Retrieval” later in this chapter, we encourage you to document all your decisions and products, and documentation needs to be maintained in an organized framework. You may prefer to use traditional methods of searching, retrieving, and storing information. For example, you may retrieve a journal article, print or photocopy it, and write notes in the margin. If you do this, you will still need to develop a cataloging system that enables you to find a particular article (e.g., alphabetical filing by last name of the first author).
Increasingly, journal articles are retrieved as portable document files (pdf) and read online using Adobe software, which permits you to highlight text passages and enter marginal comments. If this is your approach, you should create a folder on your computer or in the cloud to store these articles, naming each file in a manner that will allow you to easily locate it. For example, here is how we named the file storing
the previously mentioned Gleason et al. (2018) literature review: Gleason2018JCNAtrialFibSymptoms.pdf. This file name indicates the last name of the first author, year of publication, an abbreviation for the
journal (JCN = Journal of Cardiovascular Nursing), and a brief phrase about the topic. This system would result in a document folder with articles listed alphabetically by the first authors’ last names. You may opt to use reference management software that will help you to stay organized—as well as help you retrieve articles, maintain a reference library and notes, insert citations into papers, and create a bibliography when you write up your review. Popular reference management software that can be used with either Windows for PCs or Macs includes EndNote (free for the Basic version), Mendelay (also free), and RefWorks. Many other reference management software packages are available (for example, see
It is wise to think ahead about the various components of your literature review effort and to have a plan for how to organize them—most likely this will involve the creation of various file folders that will be
stored on your computer or in the cloud. For example, if you are not using reference management software, you should create a master folder (e.g., labeled “Pain_Management_Children”), with multiple
subfolders. For example, one subfolder could store the source documents (e.g., the pdf journal article files), another could store documentation of your search strategy and results, and another subfolder could
save drafts of your actual literature review. Another organizational tool—one that is essential for a systematic review—is a flow chart that documents your progress in identifying, retrieving, screening, and selecting source materials. Figure 5.2 presents an
example of such a flow chart with fictitious numbers (n = ) shown in each box. This figure shows that the reviewer started with 400 possible source documents, of which only 15 were used in the final literature
FIGURE 5.2 Example of a flow chart documenting literature search progress.
Locating Relevant Literature for a Research Review
As shown in Figure 5.1, an early step in a literature review is devising a strategy to locate relevant studies. The ability to locate research documents on a topic is an important skill that requires adaptability. Sophisticated new search strategies and tools are being introduced regularly. We urge you to consult with librarians, colleagues, or faculty for suggestions. Reference librarians in health libraries are especially valuable and often serve on teams conducting systematic reviews.
Formulating a Search Strategy
There are many ways to search for research evidence. Searching is inevitably an iterative process that evolves as new “leads” are discovered based on information you have already retrieved.
Search Strategy Options Cooper (2017) has identified several search strategies, one of which we describe in some detail in this chapter: searching for references in bibliographic databases. Database searches, which can be done efficiently
from computers and tablets, are likely to yield the largest number of research references—indeed, sometimes the yield can be overwhelming. Databases are searched primarily for key variables (e.g., pain management) but can also be searched for the names of researchers who have played a key role in a field. Another approach, called the ancestry approach (also called snowballing, footnote chasing, or pearl growing), involves using references cited in recent relevant studies to track down earlier research on the same topic
(the “ancestors”). This is an ongoing process that can be used to not only identify earlier relevant studies, but also to discover new search terms for subsequent electronic searches. A third method, the descendancy approach, is to find a pivotal early study and to search forward in citation indexes to find more recent studies (“descendants”) that cited the key study. Other strategies exist for
tracking down what is called the grey literature, which refers to studies with more limited distribution, such as conference papers, unpublished reports, and so on. We describe these strategies in Chapter 30 on
systematic reviews. If your intent is to “own” the literature, then you will likely want to adopt many of these strategies.
TIP You may be tempted to begin a literature search through an Internet search engine, such as Google, Yahoo, or Bing. Such a search is likely to yield a lot of “hits” on your topic but is unlikely to give you
full bibliographic information on relevant research. However, such searches can provide useful leads for search terms. Also, an Internet search may be the appropriate route for finding answers to secondary questions, such as: How many children are hospitalized annually? This information is more likely to be found in government reports, which are available online, than in research articles.
Eligibility Criteria Specifications
Search plans also involve decisions about the criteria that would make a study eligible for your review. These decisions need to be explicit to guide your search of bibliographic databases. Search limits are most often managed in databases through the use of filters (or limiters in some bibliographic software).
If you are not multilingual, you may need to constrain your search to studies written in your own language. You may want to limit your search to studies conducted within a certain time frame (e.g., within the past 15 years). You may also want to exclude studies with certain types of participants. For instance, in our example of a literature search about nurses’ management of children’s pain, we might want to exclude
studies in which the children were neonates. Constraining your search might help you to avoid irrelevant material but be cautious about putting too many restrictions on your search, especially initially. You can always make decisions to exclude studies at a
later point.
TIP Be sure not to limit your search to articles exclusively in the nursing literature (e.g., in the nursing subset of records in the database called PubMed). Researchers in many disciplines engage in research
relevant to nursing. Also, many nurse researchers publish in nonnursing journals, increasingly as members of interprofessional teams. Moreover, in some databases (e.g., PubMed), some journals with many
articles contributed by nurse researchers are not coded as being in the nursing subset (e.g., Qualitative Health Research, Birth), whereas some journals that are in the nursing subset have articles mostly not written by nurse authors (e.g., Journal of Wound Care).
Identifying Keywords Reviewers seeking articles for their reviews begin with a set of search terms, often called keywords. Thus, an important early task is to identify and make a written list of the keywords that will be used to search
bibliographic databases. The keyword list will be augmented as your search proceeds. Traditionally, the keywords are your main research variables. Many researchers use the PICO formulation (population, intervention/influence, comparison, outcome) discussed in Chapter 2 as keywords for a
literature search, although this may not always be the best strategy for systematic reviews (See Chapter 30).
In developing a list of keywords, it is important to include synonyms and to think broadly about related terms. For example, if we were searching for articles on teenage smoking, you should consider other terms
for teenage (e.g., adolescent, children, youth) and for smoking (e.g., tobacco, cigarettes). The use of a thesaurus (available in word processing software) for identifying synonyms is recommended—but take note of keywords specified by researchers themselves in articles you locate.
Searching Bibliographic Databases Reviewers typically begin by searching bibliographic databases that can be accessed by computer. The databases contain entries for millions of journal articles, and the articles are coded by professional indexers
to facilitate retrieval. For example, articles may be coded for language used (e.g., English), subject matter (e.g., pain), journal subset (e.g., nursing), and so on. Some databases can be accessed free of charge (e.g., PubMed, Google Scholar), whereas others are sold commercially—but they are often available through hospital or university libraries. Most database programs are user-friendly, offering menu-driven systems with on-screen support so that retrieval can proceed with minimal instruction. Getting Started With a Bibliographic Database Before searching an electronic database, you should become familiar with the features of the software used to access the database. The software gives you options for limiting your search, combining the results of
two searches, saving your search, and sending you notifications of new citations relevant to your search. Most programs have tutorials that can improve the efficiency and effectiveness of your search.
In most databases, there are two major strategies for searching. One method is to search for standardized subject headings (subject codes) that are assigned by indexers (usually professionals with Master’s degrees or higher in relevant disciplines). The subject headings differ from one database to another. It is useful to learn about the relevant subject codes because they offer a path to retrieving articles that use different words to describe the same concept. Another major advantage is that indexers code the articles based on a reading of the entire article (not just the abstract), and they code for meaning and not just words. Subject codes for databases can be located in the database’s thesaurus or reference tools. An alternative strategy is to enter your own keywords into a search field. Such a search is an important supplement to searching using the database’s controlled vocabulary because indexers are not infallible. However, such keyword searches are limited to searching for words in the article’s title or abstract (not in the full text), and so if concepts are not mentioned in the title or abstract, the article will not be retrieved. Most bibliographic software has automatic term mapping capabilities. Mapping is a feature that facilitates a search using your own keywords. The software translates (“maps”) the keywords you enter into the most plausible subject codes. Nevertheless, it is important to undertake both a keyword search and a subject code search because they yield overlapping but nonidentical results. General Database Search Features
Some features of an electronic search are similar across databases. One feature is the use of Boolean operators to expand or delimit a search. Three widely used Boolean operators are AND, OR, and NOT (in all caps for some databases). The operator AND delimits a search. If we searched for pain AND child, the software would retrieve only records that have both terms. The operator OR expands the search: pain OR child
could be used in a search to retrieve records with either term. Finally, NOT narrows a search: pain NOT child would retrieve all records with pain that did not include the term child. Note that when using multiple Boolean operators, they are processed from left to right. For example, the search phrase teenage AND smoking OR cigarettes would retrieve (1) records that include both teenage and smoking and (2) all records with
cigarettes, whether or not the article is about teenage smokers. Parentheses can be used to reorder the terms: teenage AND (smoking OR cigarettes). Boolean operators also can be used to combine searches for keyword terms and the last names of prominent researchers in a field, for example, teenage AND (smoking OR cigarettes) AND Kulbok (a researcher).
TIP Be extremely careful using the “NOT” operator because you run the risk of inadvertently removing relevant articles. For example, if you were searching for studies of female teenage smokers and used
“NOT male” in the search field, the software would remove any article that included both male and female participants.
Truncation symbols are another useful tool for searching databases. These symbols vary from one database to another, but their function is to expand the search. A truncation symbol (often an asterisk, *) expands a search term to include all forms of a root word. For example, a search for child* would instruct the computer to search for any word that begins with “child” such as children, childhood, or childrearing. For each database, it is important to learn what these special symbols are and how they work. For example, many databases require at least three letters at the beginning of a search term before a truncation symbol can be used. (e.g., ca* would not be allowed). Some databases (but not PubMed) allow for a wildcard symbol—often a question mark—that can be inserted into the middle of a search term to allow for alternative spellings. For example, in databases that allow wildcards, a search for behavio?r would retrieve records with either behavior or behaviour. Although truncation and wildcard symbols can sometimes be useful, they have one major drawback: in most databases, the use of special symbols turns off a software’s mapping feature. For example, a search for child* would retrieve records in which any form of “child” appeared in text fields, but it would not map any of these concepts onto the database’s subject heading codes. It may be preferable to use a Boolean
operator to list all terms of interest (e.g., child OR children), which would look for either term in a text word search of the title and abstract and would map onto the appropriate subject code. Another issue concerns phrase searching in which you want words to be kept together (e.g., blood pressure). Some bibliometric software would treat this as blood AND pressure and would search for records with
both terms somewhere in text fields, even if they are not contiguous. Quotation marks sometimes can be used to ensure that the words are searched in combination, as in “blood pressure.” PubMed recommends, however, that you do not use quotation marks until you have first tried a search without them. PubMed automatically searches for phrases during its mapping process—i.e., in searching for relevant subject heading codes. Key Electronic Databases for Nurse Researchers
Two bibliographic databases that are especially useful for nurse researchers are the Cumulative Index to Nursing and Allied Health Literature (CINAHL) and Medical Literature On-Line (MEDLINE, accessed
through PubMed), which we discuss in the next sections. We also briefly discuss Google Scholar. Other potentially useful bibliographic databases/search engines for nurses include the following:
British Nursing Index (BNI) Cochrane Central Register of Controlled Trials (CENTRAL) Cochrane Database of Systematic Reviews Database of Promoting Health Effectiveness Reviews (DoPHER) Excerpta Medica database (EMBASE) Health and Psychosocial Instruments database (HaPI) Psychology Information (PsycINFO)
In addition, the ISI Web of Knowledge and Scopus are two citation indexes for retrieving articles that cite a source article. Note that a search strategy that works well in one database does not always produce good results in another. Thus, it is important to explore strategies in each database and to understand how each database is
structured—for example, what subject codes are used, how they are organized in a hierarchy, and what special features are available.
TIP In the following sections, we provide specific information about using CINAHL and MEDLINE via PubMed. Note, however, that databases and the software through which they are accessed change
periodically, and so our instructions may not be up-to-date.
Cumulative Index to Nursing and Allied Health Literature CINAHL is an important bibliographic database: it covers references to virtually all English-language nursing and allied health journals, and includes books, dissertations, and selected conference proceedings in
nursing and allied health fields. There are several versions of the CINAHL database (e.g., CINAHL Plus, CINAHL Complete), each with somewhat different features relating to full text availability and journal coverage. The CINAHL database indexes material from more than 5,000 journals dating from 1981 and contains more than 6 million records. In addition to providing bibliographic information for references (i.e., author,
title, journal, year of publication, volume, and page numbers), CINAHL provides abstracts of most citations. Links to the actual article are sometimes provided. We illustrate features of CINAHL but note that some features may be different at your institution. At the outset, you might begin with a “basic search” by simply entering keywords or phrases relevant to your primary question. As you begin to enter your term into the search box, autocomplete suggestions will display, and you can click on the one that is the best match. In the basic search screen, you can limit your search in a number of ways, for example, limiting the records retrieved to those with certain features (e.g., only ones with abstracts; only research articles); to a specific range of publication dates (e.g., only those from 2010 to the present); or to those in specific languages (e.g., English). The search screen allows you to
expand your search by clicking an option labeled “Apply related words.” As an example, suppose we were interested in recent research on nurses’ pain management for children. If we did a keyword search for pain management, we would get about 18,000 records. Searching for pain management AND child AND nurse would bring the number down to about 400 (we did not truncate child* because this would retrieve records for some irrelevant terms associated with pain, such as childhood). We
could pare the number down to about 160 by limiting the search to research articles with abstracts published since 2000. The full records for the 160 references would then be displayed on the monitor in a Search Results list. There is a “sort” option at the top of the list that allows you to sort the references based on several criteria, such as publication date, author’s last name, and relevance. From the Results list, we could place promising references into a folder for later scrutiny by clicking on a file icon in the upper right corner of each
entry. We could then save the folder, print it, or export it to reference manager software such as EndNote. An example of an abridged CINAHL record entry for a study identified through the search on the management of children’s pain is presented in Figure 5.3. The record begins with the article title, the authors’ names and affiliation, and source. The source indicates the following:
Name of the journal (Pain Management Nursing) Year and month of publication (Feb 2015) Volume (16)
Issue (1) Page numbers (40-50)
FIGURE 5.3 Example of a record from a CINAHL (Cumulative Index to Nursing and Allied Health Literature) search.
(Abstract reprinted with permission from He H.G., Klainin-Yobas P., Ang E., Sinnappan R., Pölkki T., & Wang W. (2015). Nurses’ provision of parental guidance regarding school-aged children’s postoperative pain management: A descriptive correlational study. Pain Management Nursing , 16 , 40–50.)
The record also shows the major and minor CINAHL subject headings that were coded by the indexers. Any of these headings could have been used to retrieve this reference. Note that the subject headings
include substantive codes, such as Postoperative Pain, and methodologic codes (e.g., Correlational Studies), person characteristic codes (e.g., Child), and a location code (Singapore). Next, the abstract for the study is
shown. Based on the abstract, we might be able to decide whether this reference was pertinent. Each entry shows an accession number that is the unique identifier for each record in the CINAHL database, as well as other identifying numbers. An important feature of CINAHL helps you to find other relevant references once a good one has been found. In Figure 5.3 you can see that the record offers many embedded links on which you can click. For example, you could click on any of the authors’ names to see if they published other related articles. There is also a sidebar link in each record called Times Cited in this Database (if there has been a citation), with which you could retrieve records for articles that had cited this paper (for a descendancy search). Another link is labeled Find Similar Results that suggests other relevant references.
In CINAHL, you can also explore the structure of the database’s thesaurus to get additional leads for searching. The tool bar at the top of the screen has a tab called CINAHL Headings. When you click on this tab, you can enter a term of interest in the Browse field and select one of three options: Term Begins With, Term Contains, or Relevancy Ranked (which is the default). For example, if we entered pain management and then
clicked on Browse, we would be shown the major subject headings relating to pain management; we could then search the database for any of the listed subject codes.
TIP Note that the keywords we used to illustrate this simplified search (pain management, child, nurse) would not be adequate for a comprehensive retrieval of studies relevant to our review question. For example, we would want to search for several additional terms (e.g., pediatric).
The MEDLINE Database and PubMed
The MEDLINE database was developed by the U.S. National Library of Medicine and is widely recognized as the premier source for bibliographic coverage of the biomedical literature. MEDLINE covers about 5,600 medical, nursing, and health journals published in about 70 countries and contains more than 28 million records dating back to the mid-1940s. In 1999, abstracts of systematic reviews from the Cochrane Collaboration became available through MEDLINE. The MEDLINE database can be accessed through a commercial vendor, but it can be accessed for free through the PubMed website ( This means that anyone, anywhere in
the world with Internet access can search for journal articles, and thus PubMed is a lifelong resource. PubMed has excellent tutorials, including a 30-minute tutorial specifically for nurses (PubMed for Nurses). PubMed includes all references in the MEDLINE library plus additional references, such as those that have not yet been indexed. On the Home page of PubMed, you can launch a basic search that looks for your keywords in text fields of the record. PubMed, like CINAHL, has an autocomplete feature that offers suggestions as you begin to
enter your terms.
TIP On the PubMed home page, you can also launch a Clinical Query search, which is a particularly useful tool for searching for evidence in the context of an EBP inquiry. Supplement A to this chapter on provides guidance for undertaking such a clinical query. MEDLINE uses a controlled vocabulary called MeSH (Medical Subject Headings) to index articles. Indexers assign as many MeSH headings as appropriate to cover content and features of the article—typically 5
to 15 codes. You can learn about relevant MeSH terms by clicking on the “MeSH database” link on the Home page (under the heading More Resources). If, for example, we searched the MeSH database for “pain,” we would find that Pain is a MeSH subject heading (a definition is provided) and there are 60 related categories—for example, “Cancer pain,” “Back pain,” and “Headache.” Each category has numerous
subheadings, such as “Complications,” “Etiology,” and “Nursing.”
If you begin with a keyword search, you can see how your term mapped onto MeSH terms by looking in the right-hand panel for a section labeled Search Details. For example, if we entered “children” as our keyword in the search field of the initial screen, Search Details would show us that PubMed searched for all references that have “child” or “children” in text fields of the database record, and it also searched for all references that had been coded “child” as a subject heading because “child” is a MeSH subject heading.
If we did a PubMed search of MEDLINE similar to the one we described earlier for CINAHL, we would find that a simple search for pain management would yield about 102,000 records; a search for pain management AND child AND nurse would yield nearly 700 records. We can place restrictions on the search using filters that are shown in the left sidebar of the screen. If we limited our search to entries with
abstracts, written in English, and published in 2000 or later, the search would yield about 450 records. Thus, PubMed search yielded more references than the CINAHL search, in part because MEDLINE indexes more journals; another factor, however, is that in PubMed we could not limit the search to research articles because PubMed does not have a generic category that distinguishes research articles from nonresearch
TIP Here are the Search Details (the strategy and syntax) for the PubMed search just described: (“pain management”[MeSH Terms] OR (“pain”[All Fields] AND “management”[All Fields]) OR “pain management”[All Fields]) AND (“child”[MeSH Terms] OR “child”[All Fields]) AND (“nurses”[MeSH Terms] OR “nurses”[All Fields] OR “nurse”[All Fields]) AND (hasabstract[text] AND
(“2000/01/01”[PDAT]: “3000/12/31”[PDAT]) AND English[lang])
From the Search Results page, we would then click on links to the citations that suggest a relevant article; this would bring up a new screen that provides the abstract for the article and further details. Figure 5.4
shows the full citation and abstract for the same study we located earlier in CINAHL. Beneath the abstract, the display presents the MeSH terms that were indexed for this study. (Those marked with an asterisk, such as Pain Management/nursing, are MeSH subject headings that are a major focus of the article). As you can see, the MeSH terms are different than the subject headings for the same article in CINAHL. As with CINAHL, you can click on highlighted record entries (authors’ names and MeSH terms) for possible new leads.
FIGURE 5.4 Example of a record from a PubMed search.
(Abstract reprinted with permission from He H.G., Klainin-Yobas P., Ang E., Sinnappan R., Pölkki T., & Wang W. (2015). Nurses’ provision of parental guidance regarding school-aged children’s postoperative pain management: A descriptive correlational study. Pain Management Nursing , 16 , 40–50.)
In the right panel of the screen for specific PubMed records, there is a list of Similar Articles, which is a useful feature once you have found a study that is a good exemplar of what you are seeking. Further down in
the right panel, PubMed provides a list of any articles in the PubMed Central database that had cited this study. PubMed Central is a repository for full-text articles, so you could immediately download any of the articles that appeared in this list. You can also save articles that look pertinent to your review by clicking the button “Add to Favorites” at the top of the right panel. A useful feature of PubMed is that it provides access to new research by including citations to forthcoming articles in many journals. The records for these not yet published articles have the tag “Epub ahead of print.” McKeever et al. (2015) offer further suggestions for using PubMed for doing an exhaustive literature review.
TIP Searching for qualitative studies can pose special challenges. Wilczynski et al. (2007) described optimal search strategies for qualitative studies in the CINAHL database. Flemming and Briggs (2006) compared three alternative strategies for finding qualitative research.
Google Scholar Launched in 2004, Google Scholar (GS) has become an increasingly popular bibliographic search engine. GS includes articles in journals from scholarly publishers in all disciplines, as well as scholarly books,
technical reports, and other documents. GS is accessible free of charge over the Internet. Like other bibliographic search engines, GS allows users to search by topic, by a title, and by author and uses Boolean
operators and other search conventions. Like PubMed and CINAHL, GS has a Cited By feature for a descendancy search and a Related Articles feature to locate other sources with relevant content to an identified
article. Because of its expanded coverage of material, GS can provide access to many free full-text publications. Unlike other scholarly databases, GS does not order the retrieved references by publication date (i.e., most recent ones first). The ordering of records in GS is determined by an algorithm that puts most weight on
the number of times a reference has been cited; this in turn means that older references are usually earlier on the list. Another disadvantage of GS is that the search filters are fairly limited.
In the field of medicine, GS has generated considerable controversy, with some arguing that it is of similar utility and quality to popular medical databases (Gehanno et al., 2013), and others urging caution in depending primarily on GS (e.g., Boeker et al., 2013; Bramer et al., 2013). Some have found that for quick clinical searches, GS returns more citations than PubMed (Shariff et al., 2013). The capabilities and features of GS may improve in the years ahead, but at the moment, it may be risky to depend on GS exclusively. For a full literature review, we think it is best to combine searches using GS with searches of other databases. We note, however, that GS has been of particular interest in efforts to retrieve the so-called grey literature (Haddaway et al., 2015).
TIP For most reviews, other resources beyond bibliographic databases should be considered. Other sources include government reports, clinical trial registries (e.g.,, and records of studies
that are in progress such as in NIH RePORTER, which is a searchable database of biomedical projects funded by the U.S. government.
Screening and Gathering References
Screening references for relevance is a multiphase process. The first screen is the title of the article itself. For example, suppose our study question was the one we presented earlier: Among nurses working in
hospitals, what characteristics of the nurses or their practice settings are associated with their management of children’s pain? The PubMed search for pain management AND child AND nurse yielded about 450 references in PubMed. The title of one article identified in this search was “Nurses’ perceptions of caring for childbearing women who misuse opioids.” Based on this title, we could conclude that this article (which was
retrieved because the name of the journal in which it was published included the word “Child,” one of our keywords) would provide no evidence about factors influencing nurses’ pain management with
children. Once this initial screening is completed and the various search lists are also purged of duplicates, we would then examine the abstracts of the remaining references. When there is no abstract, or when the abstract
is ambiguous as to its relevance to your review, it is usually necessary to screen the full article. During the screening, keep in mind that some articles judged to be not relevant for your primary question may be useful for a secondary question. The next step is to retrieve the full text of references you think may have value for your review. If you are affiliated with an institution, you may have online access to most full-text articles, which you should
download and file. If you are not so fortunate, more effort will be required to obtain the full-text articles. Consulting with a librarian is a good strategy. The open-access journal movement is gaining momentum in healthcare publishing. Open-access journals provide articles free of charge online, regardless of any institutional subscriptions. Some journals have a hybrid format in which most articles are not open-access but some individual articles are designated as open-access. Bibliographic databases indicate which articles are open-access, and for these articles, the full
text can be retrieved by clicking on a link. (In PubMed, the link to click on states “Free Article” or “Free PMC article.”)
TIP We provide links to open-access articles with content relevant to each chapter of this book in the Toolkit of the accompanying Resource Manual. When an article is not available to you online, you may be able to access it by communicating with the lead author. Bibliographic databases usually provide an email address for lead authors. Another alternative
is to go to scholarly collaboration network (SCN) websites such as Research Gate or and do a search for a particular author. Authors sometimes upload articles onto their profile for access by others. If an
article has not been uploaded, these network sites provide a mechanism for you to send the author a message so that you can request an article to be sent to you directly. Documentation in Literature Retrieval
If your goal is to “own” the literature, you will be using a variety of databases, keywords, subject headings, authors’ names, and search strategies in an effort to pursue all leads. As you meander through the
complex world of research information, you will likely lose track of your efforts if you do not document your actions from the outset.
It is advisable to use word processing, spreadsheet, or reference manager software to record your search strategies and search results. You should make note of information such as names of the databases
searched; limits put on your search; specific keywords, subject headings, or authors used to direct the search; studies used to inaugurate a “Related Articles” or “descendancy” search; websites visited; links pursued; authors contacted to request further information or copies of articles not readily available; and any other information that would help you keep track of what you have done—including information
about the dates your searches were undertaken. Part of your strategy usually can be documented by saving your search history from bibliographic databases. Completing a flow chart such as the one shown in
Figure 5.2 is recommended if your goal is to publish a free-standing review. By documenting your actions, you will be able to conduct a more efficient search—that is, you will not inadvertently duplicate a strategy you have already pursued. Documentation will also help you to assess what else needs to be tried—where to go next in your search. Finally, documenting your efforts is a step in ensuring that your literature review is reproducible.
Extracting and Recording Information
Once you have a set of useful source materials, you need a strategy for making sense of the information. If a literature review is fairly simple, it may be sufficient to jot down notes about key features of the studies under review and to use these notes as the basis for the synthesis. Many literature reviews are sufficiently complex that a systematic process for extracting and recording information must be developed. In the past, researchers used paper-based data extraction forms to record
information about each reference. The use of word processing or spreadsheet software is advantageous, however, because then the forms can be easily searched and sorted. We call them data extraction forms because, in a review, the “data” are the information from each study. The data extraction forms are the critical bridge between the information in the original research reports and the synthesis of evidence by
reviewers. An approach that is gaining in popularity is the creation of two-dimensional data collection forms (matrices or evidence summary tables) in which rows are used for individual studies and columns are used to insert
relevant data about each study, such as sample characteristics, methodologic features, and results. Two-dimensional tables can provide insights into important “themes” in the data across studies.
Information to Extract
It is wise to record key information for each study in a systematic way. Regardless of what approach is used to record data, reviewers should decide in advance what information about each study is important. The key elements will vary from one review to the next, but you should have, as a goal, the creation of a file in which each study in the review is abstracted for a consistent set of features. Box 5.1 presents a list of some elements that could be considered for your data extraction forms. Not all of these elements are needed for each review, and for other reviews additional elements are likely to be needed. Although many terms in this table may not be familiar to you yet, you will learn about them in later chapters. Once you have decided on the elements you wish to use in your data extraction form, you should pilot test it with a sample of studies. If you discover later in the extraction process that other elements are needed, you would have to go back to every completed article to retrieve the new information.
TIP We encourage the use of two-dimensional data extraction forms, but if you prefer using a separate form to extract information for each study, an example is provided as a Word document in the Toolkit
for this chapter that you can adapt.
Coding the Studies for Key Variables
In systematic reviews, the review team almost always develops coding systems to support statistical analyses of study findings. Coding may not be necessary in less formal reviews, but coding certain elements can be helpful in organizing the review, and so we offer some suggestions and an example. We find it useful to code study findings for key variables (quantitative) or themes (qualitative). In our earlier example about factors affecting nurses’ management of children’s pain, nurses’ characteristics are the
independent variables and nurses’ pain management behaviors are the dependent variables. By reading the retrieved articles, we find that several characteristics have been studied—nurses’ knowledge about pain management, their nursing experience, demographic characteristics, and so on. We can assign codes to each type of factor. With regard to the dependent variable—nurses’ pain management behaviors—
some studies have focused on nurses’ pain assessment, others have examined nurses’ use of nonpharmacologic methods of pain relief, and so on. These outcome categories can also be coded. An example of a
coding scheme is presented in Box 5.2—there are eight independent variable categories and five outcome categories. The results of each study can then be coded. You can record these codes in data extraction forms, but we think it is also useful to note the codes in the margins of the articles themselves, so you can easily find the
information. Figure 5.5, which presents an excerpt from the results section of the study by He et al. (2015), shows marginal coding of key variables. In this excerpt, we see that the researchers reported that nurses’ guidance to parents about pain management (Code E) varied by the nurses’ age (Code 4), whether or not they had children of their own (Code 4), and their perceived knowledge about methods of pain relief
(Code 1).
FIGURE 5.5 Coded excerpt from the Results section of a research article: nurses’ management of children’s pain example. The codes in the margin, which here were entered as a comment on the pdf file, correspond to the codes explained in Box 5.2. Supplement B on
discusses this excerpt and why additional codes would be required.
(Excerpt reprinted with permission from He H.G., Klainin-Yobas P., Ang E., Sinnappan R., Pölkki T., & Wang W. (2015). Nurses’ provision of parental guidance regarding school-aged children’s postoperative pain management: A descriptive correlational study. Pain Management Nursing , 16 , 40–50.) When reviews are more sharply focused than the one we have used as an example, coding may not be necessary or codes that are more fine-tuned could be used. For example, if our research question focused
explicitly on nurses’ use of nonpharmacologic methods of pain relief (not about use of analgesics or pain assessment), the outcome categories might be specific nonpharmacologic approaches, such as distraction, guided imagery, music, massage, and so on. The point is to use codes to organize information in a way that facilitates retrieval and analysis. Further guidance on coding study findings is offered in Supplement B
on .
Literature Review Summary Tables As noted earlier, we recommend using two-dimensional tables (matrices) to extract and record information from the source documents because such tables directly support a thematic analysis of the retrieved
evidence. For some literature reviews—for example, in a dissertation—such tables are sometimes included directly in the written product. In other words, these tables can serve not only as a data extraction tool, but also as a display of critical information in complex reviews. As Box 5.1 suggests, the list of potential elements to be extracted from each study can be long. With two-dimensional tables for recording the extracted data, it may be advantageous to create multiple data
extraction forms, so that the information can be conveniently displayed on your computer screen without having to scroll right and left. For example, separate forms can be used for source information, methods used, results, and evaluation. Table 5.1 presents an example of one such matrix for extracting methodologic features of studies in a review. Such tables can be created in word processing or spreadsheet software. This table only shows one
illustrative entry: the study by He et al. (2015), whose CINAHL and PubMed records were shown in Figures 5.3 and 5.4. Complete evidence summary tables would have a row for each study in the review. These
tables can be electronically searched and sorted and re-sorted (e.g., by authors’ names, year of publication, level of evidence, etc.). Although we have only included one entry in this table as an illustration, if this
table listed 10 to 15 studies, we would be able to tell at a glance when and where the studies had been done, what sampling methods had been used, and so on. The scrutiny of such tables can tell us not only what has been done but can also point to gaps or problems—for example, overreliance on nurses’ self-reported pain management strategies rather than direct observation of nurses’ behaviors. Supplement B to this chapter on provides additional guidance about the use of evidence summary tables, together with more complete examples.
Critical Appraisal of the Evidence
In drawing conclusions about a body of research, reviewers must record not only factual information about studies—methodologic features and findings—but must also make judgments about the value of the
evidence. This section discusses issues relating to the appraisal of studies in the review.
TIP A distinction is sometimes made between a research critique and a critical appraisal. The latter term is favored by those focusing on the evaluation of evidence for nursing practice. The term critique is more often used when individual studies are being evaluated for their scientific merit—for example, when a manuscript is reviewed by two or more peer reviewers who make recommendations about publishing the paper, or when a person is preparing a literature review. In both critiques and appraisals, however, the point is to apply knowledge about research methods, theory, and substantive issues to draw conclusions about the validity and relevance of the findings.
Appraisals of Individual Studies As traditionally defined, a research critique is an appraisal of the strengths and weaknesses of a study. A good critique identifies areas of adequacy and inadequacy in an unbiased manner. Literature reviews mainly concern the evaluation of a body of research evidence for a literature review, but we briefly offer some advice about appraisals of individual studies. We provide support for the critical appraisal of individual studies in several ways. First, suggestions for appraising relevant aspects of a study are included at the end of each chapter. Second, it can be
illuminating to have a good model, and so Appendixes H and I of the accompanying Resource Manual include comprehensive appraisals of a quantitative and mixed methods study. Third, we offer a set of key critical appraisal questions in this chapter, in Box 5.3 (quantitative studies) and Box 5.4 (qualitative studies). The second column in these two boxes lists appraisal questions, and the
third column cross-references the more detailed appraisal guidelines in other chapters. Many questions may be too difficult for you to answer at this point, but your methodologic and appraisal skills will improve as you progress through this book. The questions in these two boxes are relevant for a rapid critical appraisal that would be conducted as part of an EBP effort, as well as for appraisals for a literature review. A few comments about these guidelines are worth noting. First, the questions in Boxes 5.3 and 5.4 mainly concern the rigor with which the researchers conducted their research. For example, there are no questions
regarding ethical issues because—while extremely important—the researchers’ handling of ethical concerns is unlikely to affect evidence quality. Second, the questions in these two boxes call for a yes or no answer (although for some, the answer may be “Yes, but…”). In all cases, the desirable answer is “yes.” A “no” suggests a possible limitation, and a
“yes” suggests a likely strength. Therefore, the more “yeses” a study gets, the stronger its evidence is likely to be. These questions can thus cumulatively suggest a global assessment: a report with 10 “yeses” is
likely to be superior to one with only 4. Our simplified guidelines have shortcomings. In particular, they are generic despite the fact that appraisals cannot use a one-size-fits-all list of questions. Some questions that are relevant to, say, clinical trials do not make sense for descriptive studies. Thus, you need to use some judgment about whether the guidelines are appropriate in your situation. Finally, there are questions in these guidelines for which there are no objective answers. Even experts sometimes disagree about what are the best methodologic strategies for a study.
TIP Students may be asked to critically appraise a study to document their mastery of research concepts. Such appraisals may be expected to be comprehensive, covering substantive, theoretical, ethical, methodologic, and interpretive aspects. The Toolkit for this chapter offers more detailed lists of questions than are included in Boxes 5.3 and 5.4 for such comprehensive appraisals.
Evaluating a Body of Research
In reviewing the literature, you would not undertake a comprehensive critical appraisal of each study—but you would need to evaluate the evidence quality in each study so that you could aggregate appraisals across studies to draw conclusions about the overall body of evidence.
In preparing a literature review for a new study, the studies under review need to be assessed with an eye to answering some broad questions. First, to what extent do the cumulative findings accurately reflect
the truth or, conversely, to what extent do methodologic flaws undermine the credibility of the evidence? Another important question to consider is: For which types of people does the evidence apply—that is, for whom is the evidence applicable? The use of literature review matrices, as described in Supplement B ( ), supports the analysis and evaluation of multiple studies. For example, if there is a column for sample size in the matrix (as in Table 5.1), one could readily see at a glance whether, for example, a lot of the evidence is from studies with small, unrepresentative samples.
TABLE 5.1 Example of an Evidence Summary Table for Methodologic Features of Relevant Studies
Author Year Country Dependent Variables (With Codes)
a Independent Variables (With Codes)
a Study Design Level of Evidence
b Sample Size, Character- istics Child Age Sampling Method Data Collection Method
He et al. 2015 Singapore E: Nurses’ provision of information regarding non- pharmacologic methods of pain management 1. Perceived knowledge of nonpharmacologic pain relief methods 2. Nursing experience 3. Demographic (age, education, having own
child) 4. Nurses’ role (staff nurse vs. more senior)
Descriptive correlational,
cross- sectional V 134 RNs in 7 pediatric wards of 2 hospitals
School– aged Convenience Questionnaire
aThe codes for the independent and dependent variables are shown in Box 5.2.
bFor this table, levels from the evidence hierarchy presented in Figure 2.2 in Chapter 2 were used—although this hierarchy is appropriate primarily for Therapy questions. Alternative hierarchies for different types of questions are described in Chapter 9.
TIP Formal systems for grading a body of evidence have been developed and will be discussed in the chapter on systematic reviews (Chapter 30).
Analyzing and Synthesizing Information
Once all the relevant studies have been retrieved, read, abstracted, and appraised, the information has to be analyzed and integrated. A literature review is not simply a summary of each previous study—it is a
synthesis that features important patterns. As previously noted, doing a literature review is similar to doing a qualitative study, particularly with respect to the analysis of the data, which in this case is the
information from the retrieved studies. In both, the focus is on identifying important themes. A thematic analysis essentially involves detecting regularities, as well as inconsistencies and “holes.” Several different types of themes can be identified, as described in Table 5.2. The reason we recommend using
literature review summary tables can be seen by reading the list of possible themes and questions: it is easier to discern patterns by reading down the columns of the matrices than by flipping through a file of
review forms or skimming through articles.
TABLE 5.2 Thematic Possibilities for a Literature Review
Nature of the Theme Questions for Thematic Analysis
Substantive What does the pattern of evidence suggest? How much evidence is there? How consistent is the body of evidence across studies? How powerful are observed effects? How persuasive is the evidence? Has the clinical significance of the findings been assessed? What gaps are there in the body
of evidence? Methodologic What types of research designs or approaches have predominated? What level of evidence is typical? What populations have been studied? Have certain groups been omitted from the research? What data collection methods have been used primarily? Are data typically of high quality? Overall, what are the methodologic strengths and deficiencies? Theoretical Which theoretical frameworks have been used—or has most research been atheoretical? How congruent are the frameworks? Generalizability/transferability To what types of people and settings do the findings apply? Do findings vary for different types of people or settings? Historical Have there been substantive, methodologic, or theoretical trends over time? Is evidence getting better? When was most research conducted? Researcher Who has been doing the research, in terms of discipline, specialty area, and nationality? Do any of the researchers have a systematic program of research devoted to this topic?
Clearly, it is not possible—even in lengthy free-standing reviews—to address all the questions in Table 5.2. Reviewers must decide which patterns to pursue. In preparing a review as part of a new study, you
would need to determine which pattern is of greatest relevance for developing an argument and providing a context for the new research.
Preparing a Written Literature Review
Writing literature reviews can be challenging, especially when voluminous information must be condensed into a few pages, as is typical for a journal article or proposal. We offer a few suggestions but acknowledge that skills in writing literature reviews develop over time. Organizing the Review
Organization is crucial in a written review. Having an outline helps to structure the narrative’s flow. If the review is complex, we recommend a written outline. The outline should list the main topics or themes to be discussed and the order of presentation. The important point is to have a plan before starting to write so that the review has a coherent progression of ideas. The goal is to structure the review in such a way
that the presentation is logical, demonstrates meaningful thematic integration, and leads to a conclusion about the state of evidence on the topic. Writing a Literature Review
It is beyond the scope of this book to offer detailed guidance on writing research reviews, but we offer a few comments on their content and style. Additional assistance is provided in books such as the ones by
Fink (2020) and Galvan and Galvan (2017).
Content of the Written Literature Review
A written research review should provide readers with an objective, organized synthesis of evidence on a topic. A review should be neither a series of quotes nor a series of abstracts. The central tasks are to digest and critically evaluate the overall evidence so as to reveal the current state of knowledge—not simply to describe what researchers have done. Although key studies may be described in some detail, it is seldom necessary to provide particulars for every reference. Studies with comparable findings often are summarized together.
Example of Grouped studies Kayser et al. (2019) summarized findings from several studies in their introduction to a study of predictors of hospital-acquired pressure injuries: “In a review of 54 studies examining risk factors of pressure
injuries…as many as 200 significant risk factors were identified (Coleman et al., 2015)…Examples of indirect risk factors studied include: incontinence, age, nutrition, diabetes, and vasopressor therapy.”
The review should demonstrate that you have considered the cumulative worth of the body of research. The review should be as objective as possible. Studies that are at odds with your hypotheses should not be omitted, and the review should not ignore a study because its findings contradict other studies. Inconsistent results should be analyzed for insights into factors that might have led to discrepancies. A literature review typically concludes with a concise summary of evidence on the topic and any gaps in the evidence. If the review is undertaken for a new study, this critical summary should demonstrate the need for the research and should clarify the basis for any hypotheses.
TIP As you progress through this book, you will acquire proficiency in critically evaluating studies. We hope you will understand the mechanics of doing a review after reading this chapter, but you probably will not be ready to write a state-of-the-art review until you have gained more skills in research methods.
Style of a Research Review
Students preparing their first written research review often struggle with stylistic issues. Students sometimes accept research findings uncritically, perhaps reflecting a common misunderstanding about the
conclusiveness of research. You should keep in mind that hypotheses cannot be proved or disproved by empirical testing, and no research question can be answered definitively in a single study. The issue is partly semantic: hypotheses are not proved; they are supported by research findings.
TIP When describing study findings, you should use phrases suggesting that results are tentative, such as the following:
Several studies have found…
Findings thus far suggest…
The study results support the hypothesis that… There appears to be good evidence that…
A related stylistic problem is the interjection of opinions into the review. The review should include opinions sparingly and should be explicit about their source. Reviewers’ opinions do not belong in a literature
review, except for assessments of study quality.
TIP The Toolkit for this chapter in the accompanying Resource Manual includes a table with examples of several stylistic flaws, and suggests possible rewordings.
Critical Appraisal of Research Literature Reviews We conclude this chapter with some advice about appraising a literature review. It is often difficult to critique a research review because the author is almost invariably more knowledgeable about the topic than
the readers. It is not usually possible to judge whether the author has included all relevant literature—although you may have suspicions if none of the citations are to recent articles. Several aspects of a review, however, are amenable to evaluation by readers who are not experts on the topic. Some suggestions for appraising written research reviews are presented in Box 5.5. (These questions could be used to review your own literature review as well.)
In assessing a literature review, the key question is whether it summarizes the current state of research evidence adequately. If the review is written as part of an original research report, an equally important question is whether the review lays a solid foundation for the new study.
Research Examples of Literature Reviews
The best way to learn about the style and organization of a research literature review is to read reviews in nursing journals. We present excerpts from two reviews that were part of the introduction to journal articles about original studies. a
Literature Review From a Quantitative Research Report
Study: Evaluation of a person-centered, theory-based intervention to promote health behaviors (Worawong et al., 2018). Statement of purpose: The purpose of this study was to test the effect of a behavioral, person-centered intervention (I) on physical activity and fruit and vegetable intake (Os) in community-living adults (P). Literature review (excerpt): “Although many researchers have tested intervention effects on health behaviors, scholars continue to be challenged to develop stronger behavioral interventions to improve
individuals’ health (Desroches et al., 2013)… Scholars have tried to promote health behaviors, such as diet and activity, by focusing individuals on the need to prevent or minimize chronic illnesses (e.g., diabetes, Estabrooks et al., 2005; Guo, Chen, Whittemore, & Whitaker, 2016; or cardiovascular disease [CVD], Edelman et al., 2006; Parra-Medina et al. 2011; Sniehotta, Scholz, & Schwarzer, 2006). These approaches rest on
the assumptions that individuals (a) value prevention highly, (b) perceive susceptibility to disease or its consequences, (c) perceive health behaviors as feasible, and (d) appreciate the connection between
behaviors and clinical outcomes. However, these assumptions are not often valid as explained below. People’s motives for health behaviors can differ from those of researchers and clinicians. People can perceive the distant risk of disease as less salient than their other life goals and may not initiate or sustain
health behaviors (Carpenter, 2010; Segar, Eccles, & Richardson, 2008; Teixeira et al., 2012). Based on a systematic review, people engage in health behaviors to meet various proximal, short-term goals more so than
to prevent a distal outcome such as disease (Rhodes, Quinlan, & Mistry, 2016). People may engage in physical activity or healthy eating to alter their moods in the short term or to look better in the long term
(Bowen, Balbuena, Baetz, & Schwartz, 2013; Lauver, Worawong, & Olsen, 2008). Thus, health behavior interventions could be strengthened by making them more patient-centered. This would involve customizing interventions on people’s choices of health behaviors and on their motives, preferences, values, goals, beliefs, characteristics, or needs (Morgan & Yoder, 2012; Rhodes et al., 2016). Patient-centered interventions can be motivational and efficacious for improving diet, activity, and clinical status in the longer term (Greaves et al., 2011; Teixeira et al., 2012). To strengthen behavioral interventions, researchers have tried to identify key components of successful dietary and activity interventions (Desroches et al., 2013; Pomerleau, Lock, Knai, & McKee, 2005). For example, interventions delivered face-to-face have been more efficacious than those without face-to-face contact on physical activity… and subsequent cardiovascular fitness… (Richards, Hillsdon, Thorogood, &
Foster, 2013), as well as on maintenance of diet and activity behaviors (Fjeldsoe, Neuhaus, Winkler, & Eakin, 2011). Researchers need to identify what other components can contribute to interventions that are
efficacious, feasible, acceptable, and cost-effective (Dombrowski, O’Carroll, & Williams, 2016; Teixeira et al., 2012).”
(Excerpt reprinted with permission from Worawong C., Borden M. J., Cooper K., Perez O., & Lauver D. (2018). Evaluation of a person-centered, theory-based intervention to promote health behaviors. Nursing Research , 67 , 6-15.)
Literature Review From a Qualitative Research Report
Study: Understanding advanced prostate cancer decision-making utilizing an interactive decision tool (Jones et al., 2018) Statement of purpose: The purposes of this study were to describe and understand the lived experiences of patients with advanced prostate cancer and their decision partners who used an interactive decision aid
(DecisionKEYS) in making informed, shared treatment decisions. Literature review (excerpt): “Prostate cancer is the most commonly diagnosed cancer in men and the second leading cause of cancer deaths in the United States. In 2016, an estimated 180,890 men will be diagnosed with prostate cancer, and approximately 26,120 men will die of the disease (American Cancer Society, 2016). In a lifetime, approximately 14% of all men will be diagnosed with prostate cancer (National Cancer Institute, 2016)…
There are numerous difficult decisions that patients with advanced prostate cancer must make, including treatment options, cost of care, and family involvement; however, over time, patients with advanced
cancer often regret some past decisions (Brom et al., 2015; Christie et al., 2015; Mahal et al., 2015). Many factors may increase the likelihood that patients will not have complete information at the time it is needed
in order to optimize decision making, for example, time constraints, forgetting to ask questions, and provider-patient miscommunication (Hillen et al., 2011; Lu et al., 2011; Shay & Lafata, 2015; Woods et al., 2013) …
Many patients with advanced prostate cancer struggle with treatment decisions… If patients and healthcare providers fail to engage in a systematic, informed, shared decision-making process (a collaborative process whereby patient and healthcare provider make a healthcare decision together, taking into account scientific/clinical evidence and the patient’s/decision partner’s values and preferences), there is a greater
chance that the patient will be dissatisfied and regretful regarding the decisions that were made (Mahal et al., 2015; Poon, 2012; Weeks et al., 2012). Moreover, decision partners may become ‘proxies’ in
interactions with healthcare providers, but they often misunderstand the patient’s informational and decision needs (Longo & Slater, 2014). Decision aids can help patients apply specific health information while actively participating in health-related decision making (O’Connor et al., 2009; Stacey et al., 2014).…Decision aids are most effective when
they are tailored, interactive, collaborative, and focused on the priorities of the individual patient (Fowler et al., 2011; Jimbo et al., 2013; Ozanne et al., 2014; Sepucha et al., 2013; Stacey et al., 2014) but interactive decision aids are rarely implemented (Jimbo et al., 2013).”
(Excerpt reprinted with permission from Jones R., Hollen P., Wenzel J., Weiss G., Song D., Sims T., & Petroni G. (2018). Understanding advanced prostate cancer decision making utilizing an interactive decision
aid. Cancer Nursing , 41 , 2-10.)
Summary Points
A research literature review is a written synthesis of evidence on a research problem. Major steps in preparing a written research review include formulating a question, devising a search strategy, developing a plan to organize and document review activities, conducting a search, screening and retrieving relevant sources, extracting key data from the sources, appraising studies, analyzing aggregated information for important themes, and writing a synthesis. Research articles are the major focus of research reviews. Information in nonresearch references—e.g., case reports, editorials—may broaden understanding of a research problem but has limited utility in summarizing research evidence. A primary source is the description of a study prepared by the researcher who conducted it; a secondary source is a description of the study written by someone else. Literature reviews should be based on primary source material.
Strategies for finding studies on a topic include the use of bibliographic databases, the ancestry approach (tracking down earlier studies cited in a reference list of a report), and the descendancy approach (using a pivotal study to search forward to subsequent studies that cited it.) Electronic searches of bibliographic databases are a key method of locating references. For nurses, the CINAHL and MEDLINE (via PubMed) databases are especially useful. Google Scholar is also a popular and free resource.
In searching a database, users can perform a keyword search that looks for searcher-specified terms in text fields of a database record (or that maps keywords onto the database’s subject codes) or they search according to subject heading codes themselves. Access to many journal articles is becoming easier through online resources, especially for articles available in an open-access format. References must be screened for relevance, and then pertinent information must be extracted for analysis. Two-dimensional evidence summary tables (matrices) facilitate the extraction and organization of data from the studies, as does a good coding scheme. A research critique (or critical appraisal) is a careful evaluation of a study’s strengths and weaknesses. Critical appraisals for a research review tend to focus on the methodologic aspects and findings of retrieved studies. The analysis of information from a literature search involves the identification of important themes—regularities (and inconsistencies) in the information. Themes can take many forms, including substantive, methodologic, and theoretic themes.
In preparing a written review, it is important to organize materials logically. The reviewers’ role is to describe study findings, the dependability of the evidence, evidence gaps, and (in the context of a new study) contributions that the new study would make.
Study Activities
Study activities are available to instructors on .
Box 5.1 Information to Consider for Data Extraction in a Literature Review
Citation Contact details of lead author Methods
Study design
Level of evidence Research tradition (qualitative) Longitudinal or cross-sectional Methods of bias control (e.g., blinding) Methods of enhancing trustworthiness (qualitative)
Participants Number of participants Power analysis information Key characteristics of the sample Age
Sex Ethnicity/race
Socioeconomic Diagnosis/disease Comorbidities Country Method of sample selection Attrition (percent dropped out)
Intervention/Independent variable(s)
Independent variable
Intervention or influence Comparison Number of (intervention) groups Specific intervention (e.g., components of a complex intervention)
Intervention fidelity
Outcomes/Dependent variables
Outcomes (or phenomena in qualitative studies) Time points for outcome data collection
For each key outcome:
Outcome definition Method of data collection (e.g., self-report, observation) Specific instrument (if relevant) Reliability, validity information
Qualitative: Summary of major themes Quantitative: for each outcome of interest Summary of results Effect size
p values Confidence intervals Subgroup analyses
Evaluation Major strengths Major weaknesses Overall quality rating Other
Theoretical framework
Funding source Key conclusions of the study authors
Broadly adapted from Table 7.3.a of the Cochrane Handbook for Systematic Reviews (Higgins & Green, 2011).
Box 5.2 Substantive Codes for a Literature Review on Factors Affecting Nurses’ Management of Children’s Pain
Codes for Nurse Characteristics Associated With Their Pain Management Behavior (Independent Variables)
1. Nurses’ pain management knowledge or specialized pain training 2. Nurses’ years of nursing experience 3. Nurses’ pain attitudes and beliefs 4. Demographic nurse factors (e.g., age, sex, education, has own children) 5. Nurses’ role/credential/status (e.g., RN, CNS, APN, NP) 6. Other nurse factors (e.g., self-efficacy, personal experience with pain) 7. Organizational factors (e.g., nurses’ workload, organizational culture) 8. Participation in interventions to improve nurses’ pain management skills
Codes for Nurses’ Pain Management Behaviors (Dependent Variables)
A. Nurses’ assessment of children’s pain B. Nurses’ pain management—general strategies C. Nurses’ use of analgesics for pain management D. Nurses’ use of nonpharmacologic methods of pain management E. Provision of guidance to parents about managing their child’s pain
Box 5.3 Guide to a Focused Critical Appraisal of Evidence Quality in a Quantitative Research Report
SECTION OF THE REPORT CRITICAL APPRAISAL QUESTIONS DETAILED GUIDELINES Method Research design Was the most rigorous design used, given the purpose of the study? What was the level of evidence for the type of question asked—and is this level the highest possible? Were suitable comparisons made to enhance interpretability? Was the number of data collection points appropriate? Was the period of follow-up (if any) adequate? Did the design minimize threats to the internal validity of the study (e.g., was randomization and blinding used, was attrition minimized)? Did the design enhance the external validity and applicability of the study results?
If there was an intervention, did it have a strong theoretical basis?
Box 9.1, page 201; Box 10.1, page 223 Box 31.1, page 720
Population and sample Was the population identified? Was the sample adequately described? Was a good sampling design used to enhance the sample’s representativeness of the population? Were sampling biases minimized? Was the sample size adequate? Was a power analysis used?
Box 13.1, page 274
Data collection and measurement Were key variables operationalized using the best possible methods (e.g., interviews, observations)? Were clinically important and patient-centered outcomes measured? Did the data collection methods yield data that were reliable, valid, and responsive?
Box 14.1, page 291; Box 15.1, page 336
If there was an intervention, was it rigorously developed and implemented? Did most participants allocated to the intervention group actually receive it? Were data collected in a manner that minimized bias?
Box 9.1, page 201; Box 10.1, page 223
Results Data analysis Were appropriate and powerful statistical methods used? Did the analysis help to control for confounding variables? Were Type I and Type II errors avoided or minimized? Were subgroup analyses undertaken to better understand the applicability of the results to different types of people?
Box 17.1, page 381 Box 18.1, page 408 Box 31.1, page 720
Findings Were the findings adequately summarized? Was information about effect size and precision of estimates (confidence intervals) presented? Were findings reported in a manner that facilitates a meta-analysis, and with sufficient information needed for EBP?
Box 17.1, page 381
Interpretation of the findings Were interpretations consistent with the study’s limitations? Were causal inferences, if any, justified? Was the clinical significance of the findings discussed? Did the report address the generalizability and applicability of the findings?
Box 21.1, page 465
Summary Assessment Despite any limitations, do the study findings appear to be valid—do you have confidence in the truth value of the results? Does the report inspire confidence about the types of people and settings for whom the evidence is applicable?
Box 5.4 Guide to a Focused Critical Appraisal of Evidence Quality in a Qualitative Research Report
Is the identified research tradition congruent with the methods used to collect and analyze data? Was an adequate amount of time spent with study participants? Was there evidence of reflexivity in the design?
Box 22.1, page 490
Sample and setting Was the group or population of interest adequately described? Were the setting and sample described in sufficient detail? Was a good method of sampling used to enhance information richness? Was the sample size adequate? Was saturation achieved?
Box 23.1, page 506
Data collection Were appropriate methods used to gather data? Were data gathered through two or more methods to achieve triangulation? Were the data of sufficient depth and richness?
Box 24.1, page 526
Procedures Do data collection and recording procedures appear appropriate? Were data collected in a manner that minimized bias?
Box 24.1, page 526
Enhancement of trustworthiness Did the researchers use effective strategies to enhance the trustworthiness/integrity of the study? Was there “thick description” of the context, participants, and findings, and was it at a sufficient level to support transferability? Do the researchers’ methodologic and clinical experience enhance confidence in the study findings and interpretations?
Box 26.1, page 580
Results Data analysis Was the data analysis strategy compatible with the research tradition and with the nature and type of data gathered? Box 25.1, page 553
Findings Were findings effectively summarized, with good use of excerpts and strong supporting arguments? Did the analysis yield an insightful, provocative, authentic, and meaningful picture of the phenomenon under investigation?
Box 25.1, page 553
Theoretical integration Were the themes or patterns logically connected to each other to form a convincing and integrated whole? Box 25.1 page 553
Interpretation of the findings Were the findings interpreted within an appropriate social or cultural context, and within the context of prior studies? Were interpretations consistent with the study’s limitations? Did the report address the transferability and applicability of the findings?
Box 25.1, page 553
Summary Assessment Do the study findings appear to be trustworthy—do you have confidence in the truth value of the results? Does the report inspire confidence about the types of people and settings for whom the evidence is applicable?
Box 5.5 Guidelines for Critically Appraising Literature Reviews
1. Is the review thorough—does it include all major studies on the topic? Does it include recent research (studies published within the previous 1-3 years)? Are studies from other related disciplines included, if appropriate? 2. Does the review rely mainly on primary source research articles? 3. Is the review merely a summary of existing work, or does it critically appraise and compare key studies? Does the review identify important trends and gaps in the literature? 4. Is the review well organized? Is the development of ideas clear? 5. Does the review use appropriate language regarding the tentativeness of prior findings? Is the review objective? Does the author paraphrase, or is there an overreliance on quotes from original sources? 6. If the review is part of a research report for a new study, does the review support the need for the study? 7. If it is a review designed to summarize evidence for clinical practice, does the review draw reasonable conclusions about practice implications?
References Cited in Chapter 5
* Boeker, M., Vach, W., & Motschall, E. (2013). Google Scholar as replacement for systematic literature searches: Good relative recall and precision are not enough. BMC Medical Research Methodology, 13, 131.
* Bramer, W. M., Giustini, D., Kramer, B., & Anderson, P. (2013). The comparative recall of Google Scholar versus PubMed in identical searches for biomedical systematic reviews. Systematic Reviews, 2, 115. Cooper, H. (2017). Research synthesis and meta-analysis: A step-by-step approach (5th ed.). Thousand Oaks, CA: Sage Publications. Fink, A. (2020). Conducting research literature reviews: From the Internet to paper (5th ed.). Thousand Oaks, CA: Sage. Flemming, K., & Briggs, M. (2006). Electronic searching to locate qualitative research: Evaluation of three strategies. Journal of Advanced Nursing, 57, 95–100. Galvan, J. L.. & Galvan, M. (2017). Writing literature reviews: A guide for students of the social and behavioral sciences (7th ed.) New York: Routledge. Garrard, J. (2017). Health sciences literature review made easy: The matrix method (5th ed.) Burlington, MA: Jones and Bartlett Publishers.
* Gehanno, J. F., Rollin, L., & Darmon, S. (2013). Is the coverage of Google Scholar enough to be used along for systematic reviews? BMC Medical Informatics and Decision Making, 13, 7. Glaser, B. (1978). Theoretical sensitivity. Mill Valley, CA: The Sociology Press. Gleason, K., Nazarian, S., & Dennison-Himmelfarb, C. (2018). Atrial fibrillation symptoms and sex, race, and psychological distress: A literature review. Journal of Cardiovascular Nursing, 33, 137–143.
* Grant, M., & Booth, A. (2009). A typology of reviews: An analysis of 14 review types and associated methodologies. Health Information and Libraries Journal, 26, 91–108. Haddaway, N., Collins, A., Coughlin, D., & Kirk, S. (2015). *The role of Google Scholar in evidence reviews and its applicability to grey literature searching. PLoS One, 10, e0138237. He, H. G., Klainin-Yobas, P., Ang, E., Sinnappan, R., Pölkki, T., & Wang W. (2015). Nurses’ provision of parental guidance regarding school-aged children’s postoperative pain management: A descriptive correlational study. Pain Management Nursing, 16, 40–50.
*Higgins, J., & Green, S., (Eds.). (2011). Cochrane handbook for systematic reviews of interventions version 5.1. Oxford: The Cochrane Collaboration.
**Jones, R., Hollen, P., Wenzel, J., Weiss, G., Song, D., Sims, T., & Petroni G. (2018). Understanding advanced prostate cancer decision making utilizing an interactive decision aid. Cancer Nursing, 41, 2–10. Kayser, S., VanGilder, C., & Lachenbruch, C. (2019). Predictors of superficial and severe hospital-acquired pressure injuries: A cross-sectional study using the International Pressure Ulcer Prevalence™ survey. International Journal of Nursing Studies, 89, 46–52.
* McKeever, L., Nguyen, V., Peterson, S., Gomez-Perez, S., & Braunschweig, C. (2015). Demystifying the search button: A comprehensive PubMed search strategy for performing an exhaustive literature review. Journal of Parenteral and Enteral Nutrition, 39, 622–635. Munhall, P. L. (2012). Nursing research: A qualitative perspective (5th ed.). Sudbury, MA: Jones & Bartlett.
* Shariff, S. Z., Bejaimal, S., Sontrop, J., Iansavichus, A., Haynes, R. B., Weir, M., & Garg, A. (2013). Retrieving clinical evidence: A comparison of PubMed and Google scholar for quick clinical searches. Journal of Medical Internet Research, 15(8), e164. Spradley, J. (1979). The ethnographic interview. New York: Holt Rinehart & Winston. Wilczynski, N., Marks, S., Haynes, R. (2007). Search strategies for identifying qualitative studies in CINAHL. Qualitative Health Research, 17, 705–710. Worawong, C., Borden, M. J., Cooper, K. Perez, O., & Lauver, D. (2018). Evaluation of a person-centered, theory-based intervention to promote health behaviors. Nursing Research, 67, 6–15.
*A link to this open-access article is provided in the Toolkit for Chapter 5 in the Resource Manual.
**This journal article is available on for this chapter.
aConsult the full research reports for references cited within these excerpted literature reviews.
Theoretical Frameworks
High-quality studies achieve a high level of conceptual integration. This means that the methods are appropriate for the research questions, the questions are consistent with existing evidence, and there is a plausible conceptual rationale for hypotheses to be tested or for the design of an intervention. For example, suppose we hypothesized that a nurse-led smoking cessation intervention would result in reduced rates of smoking among patients with cardiovascular disease. Why would we make this prediction —what is our “theory” (our theoretical rationale) about how the intervention might change people’s behavior? Do we predict that the intervention will change patients’ knowledge? motivation? sense of control over their decision-making? Our view of how the intervention would “work”—what mediates the relationship between intervention receipt and the desired outcome—should guide the design of the intervention
and the study.
In designing studies, researchers need to have a conceptualization of people’s behaviors or characteristics, and how these affect or are affected by interpersonal, environmental, or biologic forces. In high quality
research, a strong, defensible conceptualization is made explicit. This chapter discusses theoretical and conceptual contexts for nursing research problems.
Theories, Models, and Frameworks Many terms are used in connection with conceptual contexts for research, such as theories, models, frameworks, schemes, and maps. We offer guidance in distinguishing these terms but note that our definitions are not universal—indeed one confusing aspect of theory-related writings is that there is no consensus about terminology.
The term theory is used in many ways. For example, nursing instructors and students use the term to refer to classroom content, as opposed to the actual practice of performing nursing actions. In both lay and
scientific usage, the term theory connotes an abstraction.
In research, the term theory is used differently by different authors. Classically, theory refers to an abstract generalization that explains how phenomena are interrelated. In this definition, a theory embodies at
least two concepts that are related in a manner that the theory purports to explain. The purpose of traditional theories is to explain or predict phenomena. Others, however, use the term theory less restrictively to refer to a broad representation that can thoroughly describe a phenomenon. Some authors refer to this type of theory as descriptive theory. Broadly
speaking, descriptive theories are ones that describe or categorize characteristics of individuals, groups, or situations by abstracting common features observed across multiple manifestations. Descriptive theory plays an important role in qualitative studies. Qualitative researchers often strive to develop conceptualizations of phenomena that are grounded in actual observations. Descriptive theory is sometimes a precursor to predictive and explanatory theories.
Components of a Traditional Theory Concepts are the basic building blocks of a theory. Classical theories comprise a set of propositions that indicate relationships among the concepts. Relationships are denoted by such terms as “is associated with,”
“varies directly with,” or “is contingent on.” The propositions form an interrelated deductive system. Theories provide a mechanism for logically deriving new statements from the original propositions. Let us illustrate with the Theory of Planned Behavior (TPB; Ajzen, 2005), which is related to another theory called the Theory of Reasoned Action (Fishbein & Ajzen, 2010). TPB provides a framework for understanding people’s behavior and its psychological determinants. A greatly simplified construction of the TPB consists of the following propositions:
1. Behavior that is volitional is determined by people’s intention to perform that behavior. 2. Intention to perform or not perform a behavior is determined by three factors: Attitudes toward the behavior (i.e., the overall evaluation of performing the behavior) Subjective norms (i.e., perceived social pressure to perform or not perform the behavior) Perceived behavioral control (i.e., the anticipated ease or difficulty of engaging in the behavior) 3. The relative importance of the three factors in influencing intention varies across behaviors and situations.
The concepts that form the basis of the TPB include behaviors, intentions, attitudes, subjective norms, and perceived self-control. The theory, which specifies the nature of the relationship among these concepts, provides a framework for generating hypotheses relating to health behaviors. For example, we might hypothesize that compliance with a medical regimen (the behavior) could be enhanced by changing people’s attitudes
toward compliance or by increasing their sense of control. The TPB has been used as the underlying theory for studying a wide range of health decision-making behaviors and in developing health-promoting
Example using the TPB
Shi et al. (2019) used the Theory of Planned Behavior to study factors influencing patient delay in seeking treatment among people with hemorrhoids in China.
TIP Links to websites devoted to theories and conceptual models mentioned in this chapter are listed in the Toolkit of the accompanying Resource Manual for you to click on directly
Levels of Theories Theories differ in their level of generality and abstraction. The most common labels used in nursing for levels or scope of theory are grand, middle-range, and micro or practice. Grand theories or macrotheories purport to describe and explain large segments of human experience. In nursing, several grand theories offer explanations of the whole of nursing and address the nature, goals, and mission of nursing practice, as distinct from the discipline of medicine. An example of a nursing theory that has been described as a grand theory is Parse’s Humanbecoming Paradigm (Parse, 2014). Theories of relevance to researchers are often more focused than grand theories. Middle-range theories attempt to explain such phenomena as decision-making, stress, comfort, and unpleasant symptoms. Middle-range theories are more specific and more amenable to empirical testing than grand theories (Peterson & Bredow, 2017). Literally dozens of middle-range theories have been developed by or used by nurses, a few of which we briefly describe in this chapter. The least abstract level of theory is practice theory (sometimes called situation-specific theory or micro theory). Such theories are highly specific, narrow in scope, and have an action orientation. They are not always associated with research, although grounded theory studies can be a source of situation-specific theory (Peterson & Bredow, 2017). Models Conceptual models, conceptual frameworks, or conceptual schemes (we use the terms interchangeably) are a less formal means of organizing phenomena than theories. Like theories, conceptual models deal with
abstractions (concepts) that are assembled by virtue of their relevance to a common theme. Conceptual models, however, lack the deductive system of propositions that purport to explain relationships among
concepts. Conceptual models provide a perspective regarding interrelated phenomena but are more loosely structured than theories. Conceptual models can serve as springboards for generating hypotheses, but conceptual models in their entirety are not formally “tested.” (In actuality, however, the terms model and theory are sometimes used interchangeably.) The term model is often used in connection with a symbolic representation of a conceptualization. Schematic models (or conceptual maps), which are visual representations of some aspect of reality, use concepts as building blocks but with a minimal use of words. A visual or symbolic representation of a theory or conceptual framework often helps to express abstract ideas in a concise and accessible format. Schematic models are common in both qualitative and quantitative research. Concepts and linkages among them are represented through the use of boxes, arrows, or other symbols. As an example, Figure 6.1
shows Pender’s Health Promotion Model, which is a model for explaining and predicting the health-promotion component of lifestyle (Murdaugh et al., 2019). Such schematic models can be useful in succinctly
communicating linkages among concepts.
FIGURE 6.1 Pender’s Health Promotion Model.
(Retrieved from
Frameworks A framework is the overall conceptual underpinnings of a study. Not every study is based on a formal theory or conceptual model, but every study has a framework—that is, a conceptual rationale. In a study
based on a theory, the framework is a theoretical framework; in a study with roots in a conceptual model, the framework is a conceptual framework.
In most nursing studies, the framework is not an explicit theory or model, and sometimes the underlying conceptual rationale for the inquiry is not explained. Frameworks are often implicit, without being
formally described. In studies without an articulated conceptual framework, it may be difficult to figure out what the researchers thought was “going on.” Sometimes researchers fail even to adequately describe key constructs at the conceptual level. The concepts in which researchers are interested are abstractions of observable phenomena, and our world view
shapes how those concepts are defined and operationalized. Researchers should make clear the conceptual definition of their key variables, thereby providing information about the study’s framework.
In most qualitative studies, the frameworks are part of the research tradition in which the study is embedded. For example, ethnographers usually begin their work within a theory of culture. The questions that most qualitative researchers ask and the methods they use to address those questions inherently reflect certain theoretical formulations.
TIP In recent years, concept analysis has become an important enterprise among students and nurse scholars, and several methods have been proposed for undertaking a concept analysis and clarifying
conceptual definitions (e.g., Rodgers & Knafl, 2000; Walker & Avant, 2019). However, Bergdahl and Berterö (2016) have argued that concept analysis is not a suitable approach to theory development.
Example of Developing a Conceptual Definition Mollohan (2018) used Walker and Avant’s eight-step concept analysis methods to conceptually define dietary culture. Mollohan searched and analyzed 67 relevant articles identified through multiple database and proposed the following: “Dietary culture can be defined as patterned group earing behaviors that are unconsciously influenced and socially organized” (p. E2).
The Nature of Theories and Conceptual Models
Theories and conceptual models have much in common, including their origin, general nature, purposes, and role in research. In this section, we examine some characteristics of theories and conceptual models. We use the term theory in a broad sense, inclusive of conceptual models. Origin of Theories and Models
Theories, conceptual frameworks, and models are not discovered; they are invented. Theory building depends not only on observable evidence but also on the originator’s ingenuity in pulling facts together and
organizing them. Theory construction is a creative enterprise that can be undertaken by anyone who is insightful, has a firm grounding in existing evidence, and is able to knit together evidence into an intelligible pattern.
Tentative Nature of Theories and Models
Theories and conceptual models cannot be proved—they represent a theorist’s best effort to describe and explain phenomena. Today’s flourishing theory may be discredited or revised tomorrow. This may happen if new evidence or observations undermine a previously accepted theory. Or, a new theory might integrate new observations into an existing theory to yield a more parsimonious or accurate explanation
of a phenomenon. Theories and models that are not congruent with a culture’s values also may fall into disfavor over time. For example, certain psychoanalytic and structural social theories, which had broad support for decades, have come to be challenged as a result of changing views about women’s roles. Theories are deliberately invented by humans, and so they are not free from human values, which can change over time.
The Role of Theories and Models
Theories allow researchers to integrate observations and facts into an orderly scheme. The linkage of findings into a coherent structure can make a body of evidence more useful.
In addition to summarizing, theories and models can guide a researcher’s understanding of not only the what of natural phenomena but also the why of their occurrence. Theories often provide a basis for predicting phenomena. Prediction, in turn, has implications for influencing phenomena. A utilitarian theory has potential to bring about desirable changes in people’s behavior or health outcomes. Thus, theories are an important resource for developing nursing interventions. Theories and conceptual models help to stimulate research and the extension of knowledge by providing both direction and impetus. Thus, theories may serve as a springboard for advances in knowledge and the accumulation of evidence for practice. Relationship Between Theory and Research
Theory and research have a reciprocal relationship. Theories are built inductively from observations, and research evidence is an excellent source for those observations. Concepts and relationships that are validated through research become the foundation for theory development. The theory, in turn, must be tested by subjecting deductions from it (hypotheses) to systematic inquiry. Thus, research plays a dual and
continuing role in theory building. Theory guides and generates ideas for research; research assesses the worth of the theory and provides a foundation for new theories.
Conceptual Models and Theories Used in Nursing Research
Nurse researchers have used nursing and nonnursing frameworks to provide a conceptual context for their studies. This section briefly discusses several frameworks that have been found useful. Conceptual Models and Theories of Nursing
Several nurses have formulated theories and models of nursing practice. These models offer formal explanations of what nursing is and what the nursing process entails. As Fawcett and DeSanto-Madeya (2013) have noted, four concepts are central to models of nursing: human beings, environment, health, and nursing. The various models, however, define these concepts differently, link them in diverse ways, and
emphasize different relationships among them. Moreover, the models view different processes as being central to nursing. The conceptual models were not developed primarily as a base for nursing research. Most models have had more impact on nursing education and practice than on research. Nevertheless, nurse researchers have been inspired by these conceptual models in formulating research questions and hypotheses. Two nursing models that have generated particular interest as a basis for research are briefly described.
Roy’s Adaptation Model
In Roy’s Adaptation Model, humans are viewed as biopsychosocial adaptive systems who cope with environmental change through the process of adaptation (Roy & Andrews, 2009). Within the human system,
there are four subsystems: physiologic/physical, self-concept/group identity, role function, and interdependence. These subsystems constitute adaptive modes that provide mechanisms for coping with
environmental stimuli and change. Health is viewed as both a state and a process of becoming integrated and whole that reflects the mutuality of persons and environment. The goal of nursing, according to this model, is to promote client adaptation. Nursing also regulates stimuli affecting adaptation. Nursing interventions usually take the form of increasing, decreasing, modifying, removing, or maintaining internal and external stimuli that affect adaptation. Roy’s Adaptation Model has been the basis for several middle-range theories and dozens of studies.
Example Using Roy’s Adaptation Model
Frank et al. (2017) were guided by Roy’s Adaptation Model in their study of the effect of implementing a posttraumatic stress disorder screening tool for acute traumatically injured patients.
Orem’s Self-Care Deficit Nursing Theory
Some basic concepts in Orem’s Self-Care Deficit Theory include self-care, self-care deficit, and self-care agency (Orem et al., 2003). Self-care activities are what people do on their own behalf to maintain their life, health, and well-being. The ability to perform self-care is called self-care agency. Orem’s universal self-care requisites to maintain health include air, food, water, elimination, activity and rest, solitude and social
interaction, hazard prevention, and promotion of normality. Self-care deficit occurs when self-care agency is not adequate to meet a person’s self-care demands. Orem’s theory explains that patients need nursing
care when their demands for self-care outweigh their abilities.
Example Using Orem’s Theory Using Orem’s self-care deficit theory as her framework Treadwell (2017) explored depression among patients on dialysis. The researcher concluded that Orem’s theory was appropriate for identifying depression and motivation for change, and for encouraging self-care practices with hemodialysis patients.
Other Models and Middle-Range Theories Developed by Nurses
In addition to conceptual models that are designed to describe and characterize the nursing process, nurses have developed middle-range theories and models that focus on more specific phenomena of interest to nurses. Examples of middle- range theories that have been used in research include:
Beck’s (2012) Theory of Postpartum Depression; Kolcaba’s (2003) Comfort Theory;
Symptom Management Model (Dodd et al., 2001); Theory of Transitions (Meleis et al., 2000); Peplau’s (1997) Theory of Interpersonal Relations
Swanson’s (1991) Theory of Caring Reed’s (1991) Self-Transcendence Theory Pender’s Health Promotion Model (Murdaugh, Parsons, & Pender, 2019); and Mishel’s Uncertainty in Illness Theory (1990).
The latter two are briefly described here.
The Health Promotion Model Nola Pender’s Health Promotion Model (HPM) focuses on explaining health-promoting behaviors, using a wellness orientation (Murdaugh et al., 2019). According to the model (see Figure 6.1), health promotion
entails activities directed toward developing resources that maintain or enhance a person’s well-being. The model embodies several theoretical propositions that can be used to develop interventions and to gain
insight into health behaviors. For example, one HPM proposition is that people commit to behaviors from which they anticipate deriving valued benefits, and another is that perceived competence or self-efficacy
relating to a given behavior increases the likelihood of performing it. Greater perceived self-efficacy is viewed as resulting in fewer perceived barriers to a health behavior. The model also incorporates
interpersonal and situational influences on a person’s commitment to health-promoting actions.
Example Using the HPM
Eren Fidanci et al. (2017) tested the effects of an intervention based on Pender’s Health Promotion Model on the healthy life behaviors of obese children in Turkey.
Uncertainty in Illness Theory Mishel’s Uncertainty in Illness Theory (Mishel, 1990) focuses on the concept of uncertainty—a person’s inability to determine the meaning of illness-related events. According to this theory, people develop
subjective appraisals to assist them in interpreting the experience of illness and treatment. Uncertainty occurs when people are unable to recognize and categorize stimuli. Uncertainty results in the inability to
obtain a clear conception of the situation, but a situation appraised as uncertain will mobilize individuals to use their resources to adapt to the situation. Mishel’s theory as originally conceptualized was most
relevant to patients in an acute phase of illness or in a downward illness trajectory, but it has been reconceptualized to include constant uncertainty in chronic or recurrent illness. Mishel’s conceptualization of uncertainty, and her Uncertainty in Illness Scale, has been used in many nursing studies.
Example Using Uncertainty in Illness Theory
Shun et al. (2018) studied changes in patients’ degree of uncertainty in relation to levels of symptom distress and unmet care needs among patients with recurrent hepatocellular carcinoma.
Other Models and Theories Used by Nurse Researchers Many concepts of interest to nurse researchers are not unique to nursing, and so their studies are sometimes linked to frameworks that originated in other disciplines. Several of these alternative models have gained special prominence in the development of nursing interventions to promote health-enhancing behaviors. In addition to the previously described TPB, three nonnursing models or theories have often been used in nursing studies: Bandura’s Social Cognitive Theory, Prochaska’s Transtheoretical (stages of change) Model, and the Health Belief Model (HBM).
Bandura’s Social Cognitive Theory Social Cognitive Theory (Bandura, 1997, 2001), which is sometimes called self-efficacy theory, offers an explanation of human behavior using the concepts of self-efficacy and outcome expectations. Self-efficacy
concerns people’s belief in their own capacity to carry out particular behaviors (e.g., smoking cessation). Self-efficacy expectations influence the behaviors a person chooses to perform, their degree of perseverance, and the quality of the performance. Bandura identified four factors that influence a person’s cognitive appraisal of self-efficacy: (1) their own mastery experience; (2) verbal persuasion; (3) vicarious experience; and (4) physiologic and affective cues, such as pain and anxiety. The role of self-efficacy has been studied in relation to numerous health behaviors (e.g., weight control, smoking).
TIP Bandura’s self-efficacy construct is a key mediating variable in several theories discussed in this chapter. Self-efficacy has repeatedly been found to explain a significant amount of variation in people’s behaviors and to be amenable to change. As a result, self-efficacy enhancement is often a goal in interventions designed to change people’s health-related behaviors (Conn et al., 2001).
Example Using Social Cognitive Theory
Staffileno et al. (2018) evaluated a Web-based, culturally relevant lifestyle change intervention, with roots in Social Cognitive Theory, that targeted young African American women at risk for developing hypertension.
The Transtheoretical (Stages of Change) Model The Transtheoretical Model (Prochaska et al., 2002; Prochaska & Velicer, 1997) has been the basis of numerous interventions designed to change people’s problem behavior (e.g., alcohol abuse). The core construct around which other dimensions are organized is stages of change, which conceptualizes a continuum of motivational readiness to change dysfunctional behavior. The five stages of change are precontemplation, contemplation, preparation, action, and maintenance. Studies have shown that successful self-changers use different processes at each stage, suggesting the desirability of interventions that are individualized to
the person’s stage of readiness for change. The model incorporates a series of mediating variables, one of which is self-efficacy.
Example Using the Transtheoretical Model Wen et al. (2019) tested the effectiveness of a Transtheoretical Model–based intervention on the self-management of people with an ostomy.
The Health Belief Model The Health Belief Model (HBM; Becker, 1978) has become a popular framework in nursing studies focused on patient compliance and preventive healthcare practices. The model postulates that health-seeking
behavior is influenced by a person’s perception of a threat posed by a health problem and the value associated with actions aimed at reducing the threat. The major components of the HBM include perceived
susceptibility, perceived severity, perceived benefits and costs, motivation, and enabling or modifying factors. Perceived susceptibility is a person’s perception that a health problem is personally relevant or that a diagnosis is accurate. Even when one recognizes personal susceptibility, action will not occur unless the individual perceives the severity to be high enough to have serious implications. Perceived benefits are patients’ beliefs that a given treatment will cure the illness or help prevent it, and perceived barriers include the complexity, duration, and accessibility of the treatment. Motivation is the desire to comply with a
treatment. Among the modifying factors that have been identified are personality variables, patient satisfaction, and sociodemographic factors.
Example Using the HBM
Rakhshkhorshid et al. (2018) used concepts from the Health Belief Model in their study of the association of health literacy with breast cancer knowledge, perception, and screening behavior.
TIP A theoretical framework called the Theoretical Domains Framework (TDF) is being used increasingly in implementation science as a way to understand factors influencing the behaviors of healthcare professionals, as well as to facilitate the design of interventions. The TDF, which was developed by expert consensus, is a framework with 14 domains derived from 33 behavior-change theories (Michie et al., 2005).
Selecting a Theory or Model for Nursing Research As we discuss in the next section, theory can be used by qualitative and quantitative researchers in various ways. A common challenge, however, is identifying an appropriate model or theory—a task made
especially daunting because of the burgeoning number available. There are no rules for how this can be done, but there are two places to start—with the theory or model, or with the phenomenon being studied. Readings in the theoretical literature often give rise to research ideas, so it is useful to become familiar with a variety of grand and middle-range theories. Several nursing theory textbooks provide good overviews of major nurse theorists (e.g., Alligood, 2018; Butts & Rich, 2018; Morse, 2017). Resources for learning more about middle-range theories include Smith and Liehr (2018) and Peterson and Bredow (2017).
: The Supplement for this chapter on includes a table that describes 11 nursing models that have been used by researchers. The Supplement also offers references for about 100 middle-range
theories and models that have been used in nursing research, organized in broad domains (e.g., aging, mental health, pain).
If you begin with a research problem or topic and are looking for a theory, a good strategy is to examine the conceptual contexts of existing studies on a similar topic. You may find that several different theories have been used, and so the next step is to learn as much as possible about the most promising ones so that you can select a theory that is appropriate for your own study.
TIP Although it may be tempting to read about the features of a theory in a secondary source, it is best to consult a primary source and to rely on the most up-to-date reference because models are often
revised as research accumulates. However, it is also a good idea to review studies that have used the theory so that you can judge how much empirical support the theory has received and how key variables were measured. Many writers have offered advice on how to evaluate a theory for use in nursing practice and nursing research (e.g., Chinn & Kramer, 2018; Fawcett & DeSanto-Madeya, 2013; Smith & Parker, 2015). Box 6.1 presents some basic questions that can be asked in a preliminary assessment of a theory or model.
In addition to evaluating the general integrity of the model or theory, it is important to make sure that there is a proper “fit” between the theory and the research question to be studied. A critical issue is whether
the theory has done a good job of explaining, predicting, or describing constructs that are key to your research problem. A few additional questions include the following:
Has the theory been applied to similar research questions, and do the findings from prior research lend credibility to the theory’s utility for research? Are the theoretical constructs in the model or theory readily operationalized? Do instruments of adequate quality exist?
Is the theory compatible with your world view and with the world view implicit in the research question?
Testing, Using, and Developing a Theory or Framework
In this section, we describe how theory is used by qualitative and quantitative researchers. We use the term theory broadly to include conceptual models and frameworks.
Theories and Qualitative Research
Theory is almost always present, either peripherally or centrally, in studies that are embedded in a qualitative research tradition such as ethnography, phenomenology, or grounded theory. These research
traditions inherently provide an overarching framework that gives qualitative studies a theoretical grounding. However, different traditions involve theory in different ways. Sandelowski (1993) made a useful distinction between substantive theory (conceptualizations of the target phenomenon under study) and theory that reflects a conceptualization of human inquiry. Some qualitative researchers insist on an atheoretical stance vis-à-vis the phenomenon of interest, with the goal of suspending a priori conceptualizations (substantive theories) that might bias their collection and
analysis of data. For example, phenomenologists are in general committed to theoretical naiveté and explicitly try to hold preconceived views of the phenomenon in check. Nevertheless, they are guided in their
inquiries by a philosophy of phenomenology that focuses their analysis on certain aspects of a person’s lived experiences. Ethnographers typically bring a strong cultural perspective to their studies, and this perspective shapes their initial fieldwork. Ethnographers often adopt one of two cultural theories: ideational theories, which
suggest that cultural conditions stem from mental activity and ideas, or materialistic theories, which view material circumstances (e.g., resources, money, production) as the source of cultural developments. The most prominent sociologic theory in grounded theory is symbolic interaction (or interactionism), which has three underlying premises (Blumer, 1986). First, humans act toward things based on the meanings
that the things have for them. Second, the meaning of things arises out of the interaction humans have with other humans. Last, meanings are handled in, and modified through, an interpretive process in dealing with the things humans encounter. Despite having a theoretical umbrella, grounded theory researchers, like phenomenologists, attempt to hold prior substantive theory (existing knowledge and
conceptualizations about the phenomenon) in abeyance until their own substantive theory begins to emerge.
Example of a Grounded Theory Study Girardon-Perlini and Ângelo (2017) conducted a grounded theory study based on a symbolic interactionist framework to explore the experiences of rural families with relatives who had cancer. Their main
category was “Caregiving to support the family world,” which represented the family’s strategies to reconcile care for the patient and care for family life.
The use of theory in qualitative studies has been the topic of some debate. Morse (2002) called for qualitative researchers to not be “theory ignorant but theory smart” (p. 296) and to “get over” their theory phobia. Morse (2004) elaborated by noting that qualitative research does not necessarily begin with holding in check all prior knowledge of the phenomenon under study. She suggested that if the boundaries of the
concept of interest can be identified, a qualitative researcher can use these boundaries as a scaffold to inductively explore the attributes of the concept. Some qualitative nurse researchers have adopted a perspective known as critical theory as their framework. Critical theory is a paradigm that involves a critique of society and societal processes and structures, as we discuss in Chapter 22. Qualitative researchers sometimes use conceptual models of nursing as an interpretive framework, rather than as a guide for the conduct of a study. For example, some qualitative nurse researchers acknowledge
that the philosophic roots of their studies lie in conceptual models of nursing developed by Newman, Parse, or Rogers. One final note is that a systematic review of qualitative studies on a specific topic is another strategy leading to theory development. In metasyntheses (Chapter 30), qualitative studies on a topic are scrutinized to
identify essential elements. The findings from different sources are then used for theory building.
Theories and Models in Quantitative Research Quantitative researchers, like qualitative researchers, link research to theory or models in several ways. The classic approach is to test hypotheses deduced from an existing theory.
Testing an Existing Theory Theories sometimes stimulate new studies. For example, a nurse might read about Pender’s HPM (Figure 6.1), and the following type of reasoning might ensue: “If the HPM is valid, then I would expect that patients with osteoporosis who perceived the benefit of a calcium-enriched diet would be more likely to alter their eating patterns than those who perceived no benefits.” Such a conjecture can serve as a starting point for testing the model.
In testing a theory or model, quantitative researchers deduce implications (as in the preceding example) and develop hypotheses, which are predictions about the way variables would be interrelated if the theory were sound. The hypotheses are then subjected to testing through systematic data collection and analysis. The testing process involves a comparison between observed outcomes with those hypothesized. Through this process, a theory is continually subjected to potential disconfirmation. If studies repeatedly fail to disconfirm a theory, it gains support. Testing continues until pieces of evidence cannot be interpreted within the context of the theory but can be explained by a new theory that also accounts for previous findings. Theory-testing studies are most useful when researchers devise logically sound deductions from the theory, design a study that reduces the plausibility of alternative explanations for observed relationships, and
use methods that assess the theory’s validity under maximally heterogeneous situations so that potentially competing theories can be ruled out. Researchers sometimes base a new study on a theory to explain earlier descriptive findings. For example, suppose several researchers had found that nursing home residents demonstrate greater levels of noncompliance with nursing staff around bedtime than at other times. These findings shed no light on underlying causes of the problem, and so suggest no way to improve it. Explanations rooted in theories of stress might be relevant in explaining the residents’ behavior. By directly testing the theory in a study (i.e., deducing hypotheses derived from the theory), a researcher might be able to explain why bedtime is a vulnerable period for nursing home residents. Researchers sometimes combine elements from two theories as a basis for generating hypotheses. In doing this, researchers need to be thoroughly knowledgeable about both theories to see if there is an adequate
conceptual rationale for conjoining them. If underlying assumptions or conceptual definitions of key concepts are not compatible, the theories should not be combined (although perhaps elements of the two can
be used to create a new conceptual framework with its own assumptions and definitions). Tests of a theory increasingly are taking the form of testing theory-based interventions. If a theory is correct, it has implications for strategies to influence people’s health-related attitudes or behavior and hence
their health outcomes. The role of theory in the development of interventions is discussed at greater length in Chapter 28.
Example of a Theory-Based Intervention Worawong et al. (2018), whose literature review was excerpted in the previous chapter, tested the effect of a person-centered intervention on physical activity and healthy nutrition in community-living
adults. The intervention, which they called “Healthy You,” was developed using integrated concepts from two theories—Self-Regulation Theory and Self-Determination Theory.
Using a Model or Theory as an Organizing Structure Many researchers who cite a theory or model as their framework are not directly testing it, but rather using the theory as an organizational or interpretive tool. In such studies, researchers begin with a
conceptualization of nursing (or stress, health beliefs, and so on) that is consistent with that of a model developer. The researchers assume that the model used as a framework is valid and proceed to conceptualize and operationalize constructs with the model in mind. Using models in this fashion can serve a valuable organizing purpose, but such studies do not address the issue of whether the theory itself is sound.
TIP The Toolkit with the accompanying Resource Manual offers some criteria for drawing conclusions about whether researchers were truly testing a theory or using a theory as an organizational or
interpretive aid. We should note that the framework for a quantitative study need not be a formal theory such as those described in the previous section. Sometimes quantitative studies are undertaken to further explicate
constructs identified in grounded theory or other qualitative research.
Fitting a Problem to a Theory Researchers sometimes develop a set of research questions or hypotheses and subsequently try to devise a theoretical context in which to frame them. Such an approach may in some case be worthwhile, but we
caution that an after-the-fact linkage of theory to a problem does not always enhance a study. An important exception is when the researcher is struggling to make sense of findings and calls on an existing theory
to help explain or interpret them.
If it is necessary to find a relevant theory or model after a research problem is selected, the search for such a theory must begin by first conceptualizing the problem on an abstract level. For example, take the
following research question: “Do daily telephone conversations between a psychiatric nurse and a patient for 2 weeks after hospital discharge reduce rates of readmission by short-term psychiatric patients?” This
is a relatively concrete research problem, but it might profitably be viewed within the context of reinforcement theory, a theory of social support, or a theory of crisis resolution. Part of the difficulty in finding a
theory is that a single phenomenon of interest can be conceptualized in ways. Fitting a problem to a theory after-the-fact should be done with circumspection. Although having a theoretical context can enhance the meaningfulness of a study, artificially linking a problem to a theory is not
the route to scientific utility. If a conceptual model is really linked to a problem, then the design of the study, decisions about what to measure and how to measure it, and the interpretation of the findings flow
from that conceptualization.
TIP If you begin with a research question and then subsequently identify a theory or model, be willing to adapt or augment your original research problem as you gain greater understanding of the theory.
Developing a Framework in a Quantitative Study Novice researchers may think of themselves as unqualified to develop a conceptual scheme of their own. But theory development depends less on research experience than on powers of observation, grasp of a problem, and knowledge of prior research. Nothing prevents a creative and astute person from formulating an original conceptual framework for a study. The framework may not be a full-fledged theory, but it should place the issues of the study into some broader perspective. The basic intellectual process underlying theory development is induction—that is, reasoning from particular observations and facts to broader generalizations. The inductive process involves integrating what one has experienced or learned into an organized scheme. For quantitative research, the observations used in the inductive process usually are findings from other studies. When patterns of relationships among variables are derived in this fashion, one has the makings of a theory that can be put to a more rigorous test. The first step in the development of a framework, then, is to formulate a generalized scheme of
relevant concepts that is firmly grounded in the research literature. Let us use as an example a study question identified in Chapter 4, namely, What is the effect of humor on stress in patients with cancer? (See the problem statement in Box 4.2). In undertaking a literature review, we find that researchers and reviewers have suggested a myriad of complex relationships among such concepts as humor, social support, stress, coping, appraisal, immune function, and neuroendocrine function
on the one hand and various health outcomes (pain tolerance, mood, depression, health status, and eating and sleeping disturbances) on the other (e.g., Christie and Moore, 2005). While there is a fair amount of
research evidence for the existence of these relationships, it is not clear how they all fit together. Without some kind of “map” of what might be going on, it could be challenging to design a strong study—we might, for example, not measure all the key variables or we might not undertake an appropriate analysis. And, if our goal is to design a humor therapy, we might struggle in developing a strong intervention in
the absence of a framework. The conceptual map in Figure 6.2 represents an attempt to put the pieces of the puzzle together for a study involving a test of a humor intervention to improve health outcomes for patients with cancer. According
to this map, stress is affected by a cancer diagnosis and treatment both directly and indirectly, through the person’s appraisal of the situation. That appraisal, in turn, is affected by the patient’s coping skills, personality factors, and available social supports (factors that themselves are interrelated). Stress and physiological function (neuroendocrine and immunologic) have reciprocal relationships.
FIGURE 6.2 Conceptual Model of Stress and Health Outcomes in Patients with Cancer. Note that we have not yet put in a “box” for humor in Figure 6.2. How do we think humor might operate? If we see humor as having primarily a direct effect on physiologic response, we would place humor near
the bottom and draw an arrow from the box to immune and neuroendocrine function. But perhaps humor reduces stress because it helps a person cope (i.e., its effects are primarily psychological). Or maybe humor will affect the person’s appraisal of the situation. Alternatively, a nurse-initiated humor therapy might have its effect primarily because it is a form of social support. Each conceptualization has a different
implication for the design of the intervention and the study. To give but one example, if the humor therapy is viewed primarily as a form of social support, then we might want to compare our intervention with
an alternative intervention that involves the presence of a comforting nurse (another form of social support), without any special effort at including humor. This type of inductive conceptualization based on existing research is a useful means of providing theoretical grounding for a study. Of course, our research question in this example could have been addressed within the context of an existing conceptualization, such as the psychoneuroimmunology (PNI) framework of McCain et al. (2005), but hopefully our example illustrates how developing an original framework
can inform researchers’ decisions and strengthen the study. Havenga et al. (2014) offer additional tips on developing a model.
TIP We strongly encourage you to draw a conceptual map before launching an investigation based on either an existing theory or your own inductive conceptualization—even if you do not plan to formally
test the entire model or present the model in a report. Such maps are valuable heuristic devices in planning a study.
Example of Developing a New Model Hoffman et al. (2017) developed and tested a rehabilitation program for lung cancer patients. The intervention was based on their own model, which represented a synthesis of two theories, the Transitional Care Model and the Theory of Symptom Self-Management.
Critical Appraisal of Frameworks in Research Reports
It is often challenging to critically appraise the theoretical context of a published research report—or its absence—but we offer a few suggestions.
In a qualitative study in which a grounded theory is developed and presented, you probably will not be given enough information to refute the proposed theory because only evidence supporting it is presented. You can, however, assess whether the theory seems logical, whether the conceptualization is insightful, and whether the evidence in support of it is persuasive. In a phenomenologic study, you should look to see
if the researcher addressed the philosophical underpinnings of the study. The researcher should briefly discuss the philosophy of phenomenology upon which the study was based. Critiquing a theoretical framework in a quantitative report is also difficult, especially because you are not likely to be familiar with a range of relevant theories and models. Some suggestions for evaluating the
conceptual basis of a quantitative study are offered in the following discussion and in Box 6.2. The first task is to determine whether the study does, in fact, have a theoretical or conceptual framework. If there is no mention of a theory, model, or framework, you should consider whether the study’s contribution is weakened by this absence. In some cases, the research may be so pragmatic that it does not really need a theory to enhance its utility. If, however, the study involves evaluating a complex
intervention or testing hypotheses, the absence of a formally stated theoretical framework or rationale suggests conceptual fuzziness.
If the study does have an explicit framework, you must ask whether the particular framework is appropriate. You may not be able to challenge the researcher’s use of a particular theory, but you can assess whether the link between the problem and the theory is genuine. Did the researcher present a convincing rationale for the framework used? Do the hypotheses flow from the theory? Will the findings contribute
to the validation of the theory? Did the researcher interpret the findings within the context of the framework? If the answer to such questions is no, you may have grounds for criticizing the study’s framework, even though you may not be able to articulate how the conceptual basis of the study could be improved.
Research Examples
Throughout this chapter, we have mentioned studies that were based on various conceptual and theoretical models. This section presents more detailed examples of the linkages between theory and research
from the nursing research literature—one from a quantitative study and the other from a qualitative study.
Research Example From a Quantitative Study: The Health Promotion Model
Study: The relationship between religiosity and health-promoting behaviors in pregnant women (Cyphers et al., 2017) Statement of purpose: The purpose of the study was to examine the relationship between religiosity and health-promoting behaviors of women at Pregnancy Resource Centers (PRCs). Theoretical framework: The Health Promotion Model (HPM, Figure 6.1) was the guiding framework for the study: “The…HPM, a middle-range theory based on expectancy-value theory and Social Cognitive Theory, provides a holistic, multidimensional framework for exploring a person’s health-promoting behavior…Religiosity had not been previously studied with the HPM, but as religiosity can be considered a personal factor…, it was included in this research study” (p. 1430). Method: The study was conducted in eastern Pennsylvania. The researchers sampled 86 pregnant women who visited PRCs, which are community centers that offer Christian, faith-based approaches to care. Study participants completed an anonymous questionnaire in a private area of the PRC. The questionnaire was used to gather data on pregnancy intention, religiosity, health-promoting behaviors, services used
at the PRC, and demographics. Key findings: The researchers found that women who attended more classes at the centers reported more frequent health-promoting behaviors. Religiosity, attendance at religious services, and a scale measuring
“satisfaction with surrender to God” were also found to be associated with higher health-promoting behavior scores. These variables included personal factors, behavior-specific cognitions, and interpersonal
factors of Pender’s model.
Research Example From a Qualitative Study: A Grounded Theory
Study: Follow the yellow brick road: Self-management by adolescents and young adults after a stem cell transplant (Morrison et al., 2018) Statement of purpose: The purpose of the study was to understand the process adolescents and young adults use to manage their care after a stem cell transplant, and to explore self-management facilitators, barriers, processes, and behaviors. Theoretical framework: A grounded theory approach was chosen to explore the psychosocial processes that adolescents and young adults use in managing their care. The authors noted that “Grounded theory is an ideal methodology for studying complex social and psychological actions and processes. Data gathered are rich and detailed including participants’ views, actions, intentions, feelings, and life structures and
the context in which they are occurring” (p. 348). Method: Data were collected through in-depth interviews with 17 adolescents and young adults (AYA) who underwent a stem cell transplant between the ages of 13 and 25. In addition, caregivers of 13 of the AYA participants were interviewed to gain a deeper understanding of how AYA care is managed after the transplant. Interviews, which lasted about an hour, were digitally recorded and transcribed for analysis. Data collection and data analysis occurred concurrently, and data collection continued until saturation occurred. Key findings: AYA and caregiver interviewer data were integrated into one framework that was developed inductively. The metaphor of Dorothy’s journey in the Wizard of Oz was applied after theoretical brainstorming by the research team was completed. Figure 6.3 provides a graphical depiction of their framework. Key concepts include “at the mercy of transplant” (the tornado), “education and instructions”
(the yellow brick road), and “inner strength” (the Great and Powerful Oz).
FIGURE 6.3 A grounded theory of the self-management process of adolescents and young adults after a stem cell transplant. Process starts with “At the mercy of transplant” and proceeds through the cycle. Adolescents/young adults may skip setbacks and proceed to
new normal, or they may revert back to another stage and repeat the cycle. Yearn for normal, inner strength, and social support influence and are influenced by the context of SCT and self-management.
(Adapted with permission from Morrison C., Martsolf D., Borich A., Coleman K., Ramirez P., Wehrkamp N., Pai A. (2018). Follow the yellow brick road: Self-management by adolescents and young adults after a stem cell transplant. Cancer Nursing , 41 , 347–358.)
Summary Points
High-quality research requires conceptual integration, one aspect of which is having a defensible theoretical rationale for the study. Researchers demonstrate conceptual clarity by delineating a theory, model, or framework on which the study is based. A theory is a broad characterization of phenomena. As classically defined, a theory is an abstract generalization that systematically explains relationships among phenomena. Descriptive theory thoroughly describes a phenomenon. Concepts are the basic components of a theory. Classically defined theories consist of a set of propositions about the interrelationships among concepts, arranged in a logical system that permits new statements (hypotheses) to be deduced from them. Grand theories (macrotheories) attempt to describe large segments of the human experience. Middle-range theories (e.g., Pender’s HPM) are specific to certain phenomena (e.g., stress, uncertainty in illness). Concepts are also the basic elements of conceptual models, but concepts are not linked in a logically ordered deductive system. Conceptual models, like theories, provide context for nursing studies. The goal of theories and models in research is to make findings meaningful, to integrate knowledge into coherent systems, to stimulate new research, and to explain phenomena and relationships among them. Schematic models (or conceptual maps) are graphic, theory-driven representations of phenomena and their interrelationships using symbols or diagrams and a minimal use of words. A framework is the conceptual underpinning of a study, including an overall rationale and conceptual definitions of key concepts. In qualitative studies, the framework often springs from distinct research traditions.
Several conceptual models and grand theories of nursing have been developed. The concepts central to models of nursing are human beings, environment, health, and nursing. Two major conceptual models of nursing used by researchers are Roy’s Adaptation Model and Orem’s Self-Care Deficit Theory. Nonnursing models used by nurse researchers include Bandura’s Social Cognitive Theory, Prochaska’s Transtheoretical Model, and Becker’s Health Belief Model.
In some qualitative research traditions (e.g., phenomenology), the researcher avoids existing substantive theories of the phenomena under study, but there is a rich theoretical underpinning associated with the tradition itself.
Some qualitative researchers specifically seek to develop grounded theories —data-driven explanations to account for phenomena under study through inductive processes.
In the classical use of theory, researchers test hypotheses deduced from an existing theory. An emerging trend is the testing of theory-based interventions.
In both qualitative and quantitative studies, researchers sometimes use a theory or model as an organizing framework or an interpretive tool. Researchers sometimes develop a problem, design a study, and then look for a conceptual framework; such an after-the-fact selection of a framework usually is less compelling than a more systematic application of a particular theory. Even in the absence of a formal theory, quantitative researchers can inductively weave together the findings from prior studies into a conceptual scheme that provides methodologic and conceptual direction to the inquiry.
Study Activities
Study activities are available to instructors on .
Box 6.1 Some Questions for a Preliminary Assessment of a Model or Theory
Issue Questions
Theoretical clarity Are key concepts defined, and are definitions clear?
Do all concepts “fit” within the theory? Are concepts used in the theory in a manner compatible with conceptual definitions?
Are schematic models helpful, and are they compatible with the text? Are schematic models needed but not presented?
Is the theory adequately explained? Are there ambiguities?
Theoretical complexity
Is the theory sufficiently rich and detailed?
Is the theory overly complex?
Can the theory be used to explain or predict phenomena, or only to describe them?
Theoretical grounding Are the concepts identifiable in reality?
Is there a research basis for the theory? Is the basis a sound one?
Appropriateness of the theory Are the tenets of the theory compatible with nursing’s philosophy?
Are key concepts within the domain of nursing?
Importance of the theory Could research based on this theory answer critical questions for nursing?
Will testing the theory contribute to nursing’s evidence base?
General issues Are there other theories or models that would do a better job of explaining phenomena of interest?
Is the theory compatible with your world view?
Box 6.2 Guidelines for Critically Appraising Theoretical and Conceptual Frameworks in a Research Article
1. Did the report describe an explicit theoretical or conceptual framework for the study? If not, does the absence of a framework detract from the usefulness or significance of the research? 2. Did the report adequately describe the major features of the theory or model so that readers could understand the study’s conceptual basis? 3. Does the theory or model fit the research problem? Would a different framework have been more appropriate? 4. If there is an intervention, was there a cogent theoretical basis or rationale for how the intervention was expected to “work” to produce desired outcomes? 5. Was the theory or model used as a basis for generating hypotheses, or was it used as an organizational or interpretive framework? Was this appropriate? 6. Did the research problem and hypotheses (if any) naturally flow from the framework, or did the purported link between the problem and the framework seem contrived? Were deductions from the theory logical? 7. Were concepts adequately defined, and in a way that is consistent with the theory? If there was an intervention, were intervention components consistent with the theory? 8. Was the framework based on a conceptual model of nursing or on a model developed by nurses? If it was borrowed from another discipline, is there adequate justification for its use? 9. Did the framework guide the study methods? For example, was the appropriate research tradition used if the study was qualitative? If quantitative, did the operational definitions correspond to the conceptual definitions? 10. Did the researcher tie the study findings back to the framework in the Discussion section? Did the findings support or challenge the framework? Were the findings interpreted within the context of the framework?
References Cited in Chapter 6
Ajzen I. (2005). Attitudes, personality and behavior (2nd ed.). New York: McGraw Hill. Alligood M. R. (2018). Nursing theorists and their work (9th ed.). St. Louis, MO: Elsevier. Bandura A. (1997). Self-ef icacy: The exercise of control. New York: W. H. Freeman. Bandura A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1–26. Beck C. T. (2012). Exemplar: Teetering on the edge: A second grounded theory modification. In Munhall P. L. (Ed.), Nursing research: A qualitative perspective (5th ed.) (pp. 257–284). Sudbury, MA: Jones & Bartlett Learning. Becker M. (1978). The health belief model and sick role behavior. Nursing Digest, 6, 35–40. Bergdahl E., & Berterö C. (2016). Concept analysis and the building blocks of theory: Misconceptions regarding theory development. Journal of Advanced Nursing, 72, 2558–2566. Blumer H. (1986). Symbolic interactionism: Perspective and method. Berkeley: University of California Press. Butts J., & Rich K. (2018). Philosophies and theories for advanced nursing practice (3rd ed.). Burlington, MA: Jones & Bartlett. Chinn P., & Kramer M. (2018). Knowledge development in nursing: Theory and process (10th ed.). St. Louis: Mosby. Christie W., & Moore C. (2005). The impact of humor on patients with cancer. Clinical Journal of Oncology Nursing, 9, 211–218. Conn V. S., Rantz M. J., Wipke-Tevis D. D., & Maas M. L. (2001). Designing effective nursing interventions. Research in Nursing & Health, 24, 433–442.
* Cyphers N., Clements A., & Lindseth G. (2017). The relationship between religiosity and health-promoting behaviors in pregnant women. Western Journal of Nursing Research, 39, 1429–1446. Dodd M., Janson S., Facione N., Fawcett J., Froelicher E. S., Humphreys J., … Taylor D. (2001). Advancing the science of symptom management. Journal of Advanced Nursing, 33, 668–676. Eren Fidanci B., Akbayrak N., & Arslan F. (2017). Assessment of a health promotion model on obese Turkish children. Journal of Nursing Research, 25, 436–446. Fawcett J., & DeSanto-Madeya S. (2013). Contemporary nursing knowledge: Analysis and evaluation of nursing models and theories (3rd ed.). Philadelphia: F.A. Davis Company. Fishbein M., & Ajzen I. (2010). Predicting and changing behavior: The reasoned action approach. New York, NY: Psychology Press. Frank C., Schroeter K., & Shaw C. (2017). Addressing traumatic stress in the acute traumatically injured patient. Journal of Trauma Nursing, 24, 78–84.
* Girardon-Perlini N., & Ângelo M. (2017). The experience of rural families in the face of cancer. Revista Brasileira Enfermagem, 70, 550–557. Havenga Y., Poggenpoel M., & Myburgh C. (2014). Developing a model: An illustration. Nursing Science Quarterly, 27, 149–156. Hoffman A., Brintnall R., Given B., von Eye A., Jones L., & Brown J. (2017). Using perceived self-efficacy to improve fatigue and fatigability in postsurgical lung cancer patients. Cancer Nursing, 40, 1–12. Kolcaba K. (2003). Comfort theory and practice. New York: Springer Publishing Co.
* McCain N. L., Gray D. P., Walter J. M., & Robins J. (2005). Implementing a comprehensive approach to the study of health dynamics using the psychoimmunology paradigm. Advances in Nursing Science, 28, 320–332. Meleis A. I., Sawyer L. M., Im E., Hilfinger Messias D., & Schumacher K. (2000). Experiencing transitions: An emerging middle-range theory. Advances in Nursing Science, 23, 12–28.
* Michie S., Johnston M., Abraham C., Lawton R., Parker S., & Walker A. (2005). Making psychological theory useful for implementing evidence-based practice: A consensus approach. Quality & Safety in Health Care, 14, 26–33. Mishel M. H. (1990). Reconceptualization of the uncertainty in illness theory. Image: Journal of Nursing Scholarship, 22, 256–262. Mollohan E. A. (2018). Dietary culture: A concept analysis. Advances in Nursing Science, 41, E1–E12.
** Morrison C., Martsolf D., Borich A., Coleman K., Ramirez P., Wehrkamp N., … Pai A. (2018). Follow the yellow brick road: Self-management by adolescents and young adults after a stem cell transplant. Cancer Nursing, 41, 347–358. Morse J. M. (2002). Theory innocent or theory smart? Qualitative Health Research, 12, 295–296. Morse J. M. (2004). Constructing qualitatively derived theory. Qualitative Health Research, 14, 1387–1395. Morse J. M. (2017). Analyzing and conceptualizing the theoretical foundations of nursing. New York: Springer Publishing Company. Murdaugh C., Parsons M. A., & Pender N. J. (2019). Health promotion in nursing practice (8th ed.). Upper Saddle River, NJ: Pearson. Orem D., Taylor S., Renpenning K., & Eisenhandler S. (2003). Self-care theory in nursing: Selected papers of Dorothea Orem. New York: Springer. Parse R. R. (2014). The humanbecoming paradigm: A transformational worldview. Pittsburgh, PA: A Discovery International Publication. Peplau H. E. (1997). Peplau’s theory of interpersonal relations. Nursing Science Quarterly, 10, 162–167. Peterson S. J., & Bredow T. S. (2017). Middle range theories: Applications to nursing research (4th ed.). Philadelphia: Lippincott Williams & Wilkins. Prochaska J. O., Redding C. A., & Evers K. E. (2002). The transtheoretical model and stages of changes. In Lewis F. M. (Ed.), Health behavior and health education: Theory, research and practice (pp. 99–120). San Francisco: Jossey Bass. Prochaska J. O., & Velicer W. F. (1997). The transtheoretical model of health behavior change. American Journal of Health Promotion, 12, 38–48. Rakhshkhorshid M., Navaee M., Nouri N., & Safarzaii F. (2018). The association of health literacy with breast cancer knowledge, perception and screening behavior. European Journal of Breast Health, 14, 144–147. Reed P. G. (1991). Toward a nursing theory of self-transcendence. Advances in Nursing Science, 13, 64–77. Rodgers B.,, & Knafl K.,. (Eds.). (2000). Concept development in nursing: Foundations, techniques, and applications (2nd ed.). Philadelphia: Saunders. Roy C., Sr., & Andrews H. (2009). The Roy adaptation model (3rd ed.). Upper Saddle River, NJ: Pearson. Sandelowski M. (1993). Theory unmasked: The uses and guises of theory in qualitative research. Research in Nursing & Health, 16, 213–218. Shi Y., Yang D., Chen S., Wang S., Li H., Ying J., … Sun J. (2019). Factors influencing patient delay in individuals with haemorrhoids: A study based on theory of planned behavior and common sense model. Journal of Advanced Nursing, 75(5), 1018–1028. Shun S., Chou Y., Chen C., & Yang J. (2018). Change of uncertainty in illness and unmet care needs in patients with recurrent hepatocellular carcinoma during active treatment. Cancer Nursing, 41, 279–289. Smith M. J., & Liehr P. (2018). Middle-range theory for nursing (4th ed.). New York, NY: Springer Publishing Co. Smith M. C., & Parker M. (2015). Nursing theories and nursing practice (4th ed.). Philadelphia: F.A. Davis. Staffileno B., Tangney C., & Fogg L. (2018). Favorable outcomes using an eHealth approach to promote physical activity and nutrition among you African American women. Journal of Cardiovascular Disease, 33, 62–71. Swanson K. M. (1991). Empirical development of a middle-range theory of caring. Nursing Research, 40, 161–166. Treadwell A. A. (2017). Examining depression in patients on dialysis. Nephrology Nursing Journal, 44, 295–307. Walker L. O., & Avant K. C. (2019). Strategies for theory construction in nursing (6th ed.). Upper Saddle River, NJ: Prentice Hall. Wen S., Li J., Wang A., Lv M., Li H., Lu Y., & Zhang J. (2019). Effects of transtheoretical-model-based intervention on the self-management of patients with an ostomy: A randomised controlled trial. Journal of Clinical Nursing, 28(9–10), 1936–1951. Worawong C., Borden M. J., Cooper K. Perez O., & Lauver D. (2018). Evaluation of a person-centered, theory-based intervention to promote health behaviors. Nursing Research, 67, 6–15.
*A link to this open-access article is provided in the Toolkit for Chapter 6 in the Resource Manual.
**This journal article is available on for this chapter.
Planning a Nursing Study
Advance planning is required for all research. This chapter provides advice for planning qualitative and quantitative studies.
Tools and Concepts for Planning Rigorous Research
This section discusses key methodologic concepts and tools in meeting the challenges of doing rigorous research.
Inference is an integral part of doing and evaluating research. An inference is a conclusion drawn from the study evidence, taking into account the methods used to generate that evidence. Inference is the attempt
to come to conclusions based on limited information, using logical reasoning processes.
Inference is necessary because researchers use proxies that “stand in” for the things that are fundamentally of interest. A sample of participants is a proxy for an entire population. A study site is a proxy for all
relevant sites in which the phenomena of interest could unfold. A control group that does not receive an intervention is a proxy for what would happen to those receiving the intervention if they did not receive it. Researchers face the challenge of using methods that yield persuasive evidence in support of inferences they wish to make. Reliability, Validity, and Trustworthiness Researchers want their inferences to correspond with the truth. Research cannot contribute evidence to guide clinical practice if the findings are biased or fail to represent the experiences of the target group. Consumers of research need to assess the quality of a study’s evidence by evaluating the conceptual and methodologic decisions the researchers made, and those who do research must strive to make decisions
that result in high-quality evidence. Quantitative researchers use several criteria to assess the rigor of a study, sometimes referred to as its scientific merit. Two especially important criteria are reliability and validity. Reliability refers to the accuracy and consistency of information obtained in a study. The term is most often associated with the methods used to measure variables. For example, if a thermometer measured Alan’s temperature as 98.1°F
1 minute and as 102.5°F the next minute, the reliability of the thermometer would be suspect. Validity is a more complex concept that broadly concerns the soundness of the study’s evidence—whether the findings are unbiased and well grounded. Like reliability, validity is a key criterion for evaluating methods to measure variables. In this context, the validity question is whether the methods are really measuring the concepts that they purport to measure. Is a self-reported measure of depression really measuring depression? Or is it measuring something else, such as loneliness? Researchers strive for solid conceptualizations of research variables and valid methods to operationalize them. Validity is also relevant with regard to inferences about the effect of the independent variable on the dependent variable. Did a nursing intervention really bring about improvements in patients’ outcomes—or were other factors responsible for patients’ progress? Researchers make numerous methodologic decisions that influence this type of study validity. Yet another validity question concerns whether the evidence
can validly be extrapolated to people who did not participate in the study. Qualitative researchers use different criteria (and different terminology) in evaluating a study’s quality. Qualitative researchers pursue methods of enhancing the trustworthiness of the study evidence (Lincoln &
Guba, 1985). Trustworthiness encompasses several dimensions—credibility, transferability, confirmability, dependability, and authenticity—which are described in Chapter 26. Credibility, an especially important aspect of trustworthiness, is achieved to the extent that the research methods inspire confidence that the results and interpretations are truthful. Credibility can be enhanced in
various ways, but one strategy merits early discussion because it has implications for the design of all studies, including quantitative ones. Triangulation is the use of multiple sources or referents to draw
conclusions about what constitutes the truth. In a quantitative study, this might mean using multiple measures of an outcome variable to see if predicted effects are consistent. In a qualitative study, triangulation might involve trying to reveal the complexity of a phenomenon by using multiple means of data collection to converge on the truth (e.g., having in-depth discussions with participants, as well as watching their behavior in natural settings). Or, it might involve triangulating the interpretations of multiple researchers working together as a team. Nurse researchers are increasingly triangulating across paradigms—that is,
integrating both qualitative and quantitative data in a mixed methods study to enhance the validity of the conclusions (Chapter 27).
Example of Triangulation
Bower et al. (2018) conducted an exploratory study of nurses’ decision-making when they are interrupted during administration of medication in the pediatric intensive care unit (PICU). During their
fieldwork, the researchers conducted in-depth interviews with PICU nurses and made observations during medication administration. Nurse researchers need to design their studies in such a way that the reliability, validity, and trustworthiness of their studies are maximized. This book offers advice on how to do this. Bias A bias is an influence that produces a distortion or error. Bias can threaten a study’s validity and trustworthiness and is a major concern in designing a study. Bias can result from factors that need to be
considered in planning a study. These include the following:
Participants’ lack of candor. Sometimes people distort their behavior or statements—consciously or subconsciously—to present themselves in the best light. Researcher subjectivity. Investigators may distort inferences in the direction of their expectations or in line with their own experiences—or they may unintentionally communicate their expectations to participants and thereby induce biased responses. Sample imbalances. The sample itself may be biased; for example, if a researcher studying abortion attitudes included only members of right-to-life (or pro-choice) groups in the sample, the results would be distorted. Faulty methods of data collection. Inadequate methods of capturing concepts can lead to biases; for example, a flawed measure of patient satisfaction with nursing care may exaggerate or underestimate patients’ concerns.
Inadequate study design. A researcher may structure the study in such a way that an unbiased answer to the research question cannot be achieved. Flawed implementation. Even a well-designed study can sustain biases if the study protocols are not carefully implemented. A researcher’s job is to reduce or eliminate bias to the extent possible, to establish mechanisms to detect or measure it when it exists, and to take known biases into account in interpreting study findings. The job
of consumers is to scrutinize methodologic decisions to reach conclusions about whether biases undermined the study evidence. Unfortunately, bias can seldom be avoided totally because the potential for its occurrence is pervasive. Some bias is haphazard. Random bias (or random error) is essentially “noise” in the data. When error is
random, distortions are as likely to bias results in one direction as the other. Systematic bias, on the other hand, is consistent and distorts results in a single direction. For example, if a scale consistently measured
people’s weights as being 2 pounds heavier than their true weight, there would be systematic bias in the data on weight. Researchers adopt a variety of strategies to eliminate or minimize bias and strengthen study rigor. Triangulation is one such approach, the idea being that multiple sources of information or points of view can
help counterbalance biases and offer avenues to identify them. Methods that quantitative researchers use to combat bias often involve research control. Research Control Quantitative researchers usually make efforts to control aspects of the study. Research control typically involves holding constant other influences on the dependent variable so that the true relationship between
the independent and dependent variables can be understood. In other words, research control attempts to eliminate contaminating factors that might obscure the relationship between the variables of central
interest. Contaminating factors—called confounding (or extraneous) variables—can best be illustrated with an example. Suppose we were studying whether urinary incontinence (UI) affects depression. Prior evidence
suggests a link, but the question is whether UI itself (the independent variable) contributes to higher levels of depression, or whether other factors account for the relationship between UI and depression. We need
to design a study to control other determinants of depression that are also related to the independent variable, UI. One confounding variable in this situation is age. Levels of depression tend to be higher in older people; people with UI tend to be older than those without this problem. In other words, perhaps age is the real cause of higher depression in people with UI. If age is not controlled, then any observed relationship between UI and depression could be caused by UI or by age. Three possible explanations might be portrayed schematically as follows:
1. UI → depression 2. Age → UI → depression
The arrows here symbolize a causal mechanism or an influence. In Model 1, UI directly affects depression, independent of any other factors. In Model 2, UI is a mediating variable—the effect of age on depression
is mediated by UI. According to this representation, age affects depression through the effect that age has on UI. In Model 3, both age and UI have separate effects on depression and age also increases the risk of UI. Some research is specifically designed to test paths of mediation and multiple causation, but in the present example, age is extraneous to the research question. We want to design a study so that the first explanation can be tested. Age must be controlled if our goal is to explore the validity of Model 1, which posits that, no matter what a person’s age, having UI makes a person more vulnerable to depression. How can we impose such control? There are several ways (Chapter 10), but the general principle is that confounding variables must be held constant. The confounding variable must somehow be handled so that,
in the context of the study, it is not related to the independent variable or the outcome. As an example, let us say we wanted to compare the average scores on a depression scale for those with and without UI. We would want to design a study in such a way that the ages of those in the UI and non-UI groups are comparable, even though, in general, the groups are not comparable in terms of age. By exercising control over age, we would have more confidence in explaining the relationship between UI and depression. The world is complex: many variables are interrelated in complicated ways. When
studying a problem in a quantitative study, it is difficult to examine this complexity directly; researchers analyze only a few relationships at a time. The value of the evidence in quantitative studies is often related
to how well researchers controlled confounding influences. In the present example, we identified one variable (age) that could affect depression, but dozens of others might be relevant (e.g., social support, self– efficacy). Researchers need to isolate the independent and dependent variables in which they are interested and then identify confounding variables that need to be controlled. Confounding variables need to be controlled only if they are simultaneously related to both the dependent and independent variables, as explained in the Supplement to this chapter on . Research control is a critical tool for managing bias and enhancing validity in quantitative studies. Sometimes, however, too much control can introduce bias. If researchers tightly control the ways in which key
study variables are manifested, the true nature of those variables may be obscured. In studying phenomena that are poorly understood or whose dimensions have not been clarified, a qualitative approach that allows flexibility and exploration is more appropriate. Randomness
For quantitative researchers, bias reduction often involves randomness—having features of the study established by chance rather than by researcher preference. When people are selected at random to participate in the study, for example, each person in the initial pool has an equal probability of being selected—which means that there are no systematic biases in the sample’s makeup. Similarly, if participants are assigned randomly to groups that will be compared (for example intervention and “usual care” groups), then there would be no systematic biases in the groups’ composition. Randomness is a compelling method of controlling confounding variables and reducing bias.
Example of Randomness Van der Meulen et al. (2018) tested a protocol that involved screening with the Distress Thermometer and Problem List for patients with head and neck cancer. A total of 110 patients were assigned, at
random, to either the Distress Thermometer intervention or to usual care. The two groups were then compared in terms of cancer worry, depressive symptoms, and quality of life. Qualitative researchers almost never consider randomness a desirable tool. Qualitative researchers tend to use information obtained early in the study in a purposeful (nonrandom) fashion to guide their inquiry
and to pursue information-rich sources that can help them expand or refine their conceptualizations. Researchers’ judgments are viewed as indispensable vehicles for uncovering the complexities of phenomena of interest. Reflexivity Qualitative researchers do not use research control or randomness, but they are as interested as quantitative researchers in discovering the truth about human experience. Qualitative researchers often rely on
reflexivity to guard against personal bias in making judgments. Reflexivity is the process of reflecting critically on the self and of analyzing and recording personal values that could affect data collection and
interpretation. Schwandt (2007) has described reflexivity as having two aspects. The first concerns an acknowledgment that the researcher is part of the setting or context under study. The second involves self-reflection about one’s own biases, preferences, and fears about the research. Qualitative researchers are encouraged to explore these issues, to be reflexive about decisions made during the inquiry, and to note their reflexive
thoughts in personal journals. As Patton (2015) noted, “To excel in qualitative inquiry requires keen and astute self-awareness” (p. 71). Reflexivity can be a useful tool in quantitative as well as qualitative research. Self-awareness and introspection can enhance the quality of any study.
Example of Reflexivity Currie and Szabo (2019) explored parents’ perspectives on caring for a child with a rare disease. Reflexivity played an important role in the analysis and interpretation of their interview data with 15 parents: “Data were analyzed considering reflexivity throughout the process…The interpretation is a process of cocreation between the researcher and the participant through reinterpretation and reflection” (p. 97).
Generalizability and Transferability Nurses increasingly rely on evidence from research in their clinical practice. Evidence-based practice is based on the assumption that study findings are not unique to the people, places, or circumstances of the original research (Polit & Beck, 2010). Generalizability is a criterion used in quantitative studies to assess the extent to which findings can be applied to people and settings beyond those used in a study. How do researchers enhance the generalizability of a study? First and foremost, they must design studies strong in reliability and validity. There is no point in wondering whether results are generalizable if they are not accurate or valid. In
selecting participants, researchers must also give thought to the types of people to whom the results might be generalized—and then select participants in such a way that the sample reflects the population of
interest. If a study is intended to have implications for male and female patients, then men and women should be included as participants. Several chapters in this book describe strategies for enhancing
generalizability. Qualitative researchers do not specifically aim for generalizability, but they do want to generate knowledge that could be useful in other situations. Lincoln and Guba (1985), in their influential book on
naturalistic inquiry, discussed the concept of transferability, the extent to which qualitative findings can be transferred to other settings, as an aspect of a study’s trustworthiness. An important mechanism for promoting transferability is the amount of rich descriptive information qualitative researchers provide about study contexts. Transferability in qualitative research is discussed in Chapter 26.
TIP Researchers are increasingly paying attention to the applicability of their findings—that is, the extent to which findings can be applied to individuals or small subgroups. We discuss this issue at length in Chapter 31.
Stakeholder Engagement There is growing agreement within the healthcare community that greater stakeholder engagement is needed in all phases of research, beginning in the planning phase—or even earlier, during the identification
of a research problem. Proponents of stakeholder involvement during the planning and implementation of health research argue that it enhances the relevance and transparency of the research and accelerates the adoption of research evidence in practice.
TIP In Europe, advocates often use the term patient and public involvement (PPI). In the United States, the Patient-Centered Outcomes Research Institute (PCORI) was established in 2010 to fund research
that can help patients make better healthcare choices, and patients play a role in guiding the research agenda. Although patients have been identified as key stakeholders, researchers can consider involving others in planning a study. Concannon et al. (2012) developed a taxonomy to guide researchers in this new era of stakeholder-engaged research and proposed this definition of “engagement” of stakeholders: “A bi-directional relationship between the stakeholder and the researcher that results in informed decision-making
about the selection, conduct, and use of research” (p. 986). They created a framework called the 7Ps to aid in the identification of stakeholders: Patients and the public; providers (e.g., nurses, physicians); purchasers; payers; policy makers; product makers; and principal investigators. Researchers need to identify key stakeholders and determine how best to involve them in the planning process. Overview of Research Design Features
A study’s research design spells out the basic strategies that researchers adopt to develop evidence that is accurate and interpretable. The research design incorporates some of the most important methodologic decisions that researchers make, particularly in quantitative studies. Table 8.1 describes seven design features that need to be considered in planning a quantitative study; several are also pertinent in qualitative studies. These features include:
whether or not there will be an intervention; how confounding variables will be controlled; whether blinding will be used to avoid biases; what the relative timing for collecting data on dependent and independent variables will be; what types of comparisons will be made to enhance interpretability; what the location of the study will be; and what timeframes will be adopted.
TABLE 8.1 Key Research Design Features in Quantitative Studies
Feature Key Questions Design Options
Intervention Will there be an intervention? What will the intervention entail? What specific design will be used?
Experimental (randomized controlled trial), quasi-experimental, nonexperimental (observational) design
Control over confounding variables How will confounding variables be controlled? Which confounding variables will be controlled? Matching, homogeneity, blocking, crossover, randomization, statistical control
Blinding (masking) From whom will critical information be withheld to avoid bias? Open versus closed study; single-blind, double-blind (with blinded groups specified) Relative timing When will information on independent and dependent variables be collected—looking backward or forward? Retrospective, prospective design Comparisons What type of comparisons will be made to illuminate key processes or relationships? What is the nature of the comparison? Within-subject design, between-subject design, mixed design, external comparisons Location Where will the study take place? Single site versus multisite; in the field vs. controlled setting Timeframes How often will data be collected? When, relative to other events, will data be collected? Cross-sectional, longitudinal design; repeated measures design
Note: Many terms in this table are explained in subsequent chapters. This section discusses the last three features because they are relevant in planning both qualitative and quantitative studies. Chapters 9 and 10 elaborate on the first four.
TIP Design decisions affect the integrity of your findings. These decisions may influence whether you receive funding (if you seek financial support) or whether your findings will be published (if you
submit to a journal). Therefore, a great deal of care and thought should go into these decisions during the planning phase.
In most quantitative (and some qualitative) studies, researchers incorporate comparisons into their design to provide a context for interpreting results. Most quantitative research questions involve either an
explicit or an implicit comparison. For example, if our research question asks what is the effect of massage on anxiety in hospitalized patients, the implied comparison is massage versus no massage, which is the
independent variable. Researchers can structure their studies to make various types of comparison, the most common of which are as follows:
1. Comparison between two or more groups. For example, if we were studying the emotional consequences of having a mastectomy, we might compare the emotional status of women who had a mastectomy with that of women with breast cancer who did not have a mastectomy. Or, we might compare those receiving a special intervention with those receiving “usual care.” In a qualitative study, we might compare mothers and fathers with respect to their experience of having a child diagnosed with leukemia. 2. Comparison of one group’s status at two or more points in time. For example, we might want to compare patients’ levels of stress before and after introducing a procedure to reduce preoperative stress. Or we might want to compare coping processes among caregivers of patients with AIDS early and later in the caregiving experience. 3. Comparison of one group’s status under dif erent circumstances. For example, we might compare people’s heart rates during two different types of exercise. 4. Comparison based on relative rankings. If, for example, we hypothesized a relationship between the pain level and degree of hopefulness in patients with cancer, we would be asking whether those with high levels of pain felt less hopeful than those with low levels of pain. This
research question involves a comparison of those with different rankings—higher versus lower—on both variables. 5. Comparison with external data. Researchers may compare their results with those from other studies or with norms (standards from a large and representative sample). This type of comparison often supplements rather than replaces other comparisons. In quantitative studies,
this approach is useful primarily when the dependent variable is measured with a reliable, well-accepted method (e.g., blood pressure readings, scores on a respected measure of depression).
Example of Using Comparative Data From External Sources Dias et al. (2018) studied the health status of bereaved parents during the first 6 months after their child’s death. They used measures of health and well-being for which national comparative data were available, which enabled them to compare their participants’ outcomes with national norms for adults in the United States.
Research designs for quantitative studies can be categorized based on the type of comparisons that are made. Studies that compare different people (as in examples 1 and 4) are between-subjects designs. Sometimes, however, it is preferable to make comparisons for the same participants at different times or under difference circumstances, as in examples 2 and 3. Such designs are within-subjects designs. When
two or more groups of people are followed over time, the design is sometimes called a mixed design because comparisons can be both within groups over time or between different groups at a given point in
time. Comparisons provide a context for interpreting the findings. In example 1 regarding the emotional status of women who had a mastectomy, it would be difficult to know whether the women’s emotional state was worrisome without comparing it with that of others—or comparing it to their state at an earlier time (for example, prior to diagnosis). In designing a study, quantitative researchers choose comparisons that will best illuminate answers to the research question. Qualitative researchers sometimes plan to make comparisons when they undertake an in-depth study, but comparisons are rarely their primary focus. Nevertheless, patterns emerging in the data often suggest
that certain comparisons have rich descriptive value.
TIP Try not to make design decisions single-handedly. Seek the advice of faculty or colleagues; patient input may also be desirable. Once you have made design decisions, consider writing out a rationale for your choices and sharing it with others to see if they can suggest improvements. A worksheet for documenting design decisions is available in the Toolkit of the accompanying Resource Manual.
Research Location An important planning task is to identify sites for the study. In some situations the study site is a “given,” as might be the case for a clinical study conducted in a hospital or institution with which researchers are affiliated, but in other studies the identification of an appropriate site involves considerable effort. The closer the setting is to the “real world,” the more relevant the evidence is likely to be to clinical practice
(Chapter 31). Planning the study location involves two types of activities—selecting the site or sites and gaining access to them. Although some of the issues we discuss here are of particular relevance to qualitative researchers working in the field, many quantitative studies also need to attend to these matters in planning a project, especially in intervention studies.
Site Selection The primary consideration in site selection is whether the site has people with the behaviors, experiences, or characteristics of interest. The site must also have a sufficient number of these kinds of people and
adequate diversity or mix of people to achieve research goals. In addition, the site must be one in which access to participants will be granted. Both methodologic goals (e.g., ability to impose needed controls) and
ethical requirements (e.g., ability to ensure privacy and confidentiality) need to be achieved in the chosen site. Researchers sometimes must decide how many sites to include. Having multiple sites is advantageous for enhancing the generalizability of the study findings, but multisite studies are complex and challenging. Multiple sites are a good strategy when several coinvestigators from different institutions are working together on a project. Site visits to potential sites and clinical fieldwork are useful to assess the “fit” between what the researcher needs and what the site has to offer. During site visits, the researcher makes observations and converses with key gatekeepers or stakeholders in the site to better understand its characteristics and constraints. Buckwalter et al. (2009) have noted particular issues of concern when working in sites that are “unstable”
research environments, such as critical care units or long-term care facilities. Gaining Entrée Researchers must gain entrée into the sites deemed suitable for the study. If the site is an entire community, a multitiered effort of gaining acceptance from gatekeepers may be needed. For example, it may be necessary to enlist the cooperation first of community leaders and subsequently of administrators and staff in specific institutions (e.g., domestic violence organizations) or leaders of specific groups (e.g., support groups). Because establishing trust is a central issue, gaining entrée requires strong interpersonal skills, as well as familiarity with the site’s customs and language. Researchers’ ability to gain the gatekeepers’ trust can best occur if researchers are candid about research requirements and express genuine interest in and concern for people in the site. Gatekeepers are most likely to be cooperative if they believe that there will be direct benefits to them or their constituents.
Information to help gatekeepers make a decision about granting access usually should be put in writing, even if the negotiation takes place in person. An information sheet should cover the following points: (1)
the purpose and significance of the research; (2) why the site was chosen; (3) what the research would entail (e.g., study timeframes, how much disruption there might be, what resources are required); (4) how
ethical guidelines would be maintained, including how results would be reported; and (5) what the gatekeeper or others at the site have to gain from cooperating in the study. Figure 8.1 presents an example of a
letter of inquiry for gaining entrée into a facility.
FIGURE 8.1 Sample letter of inquiry for gaining entrée into a research site (fictitious).
Gaining entrée may be an ongoing process of establishing relationships and rapport with people at the site, including prospective informants. The process might involve progressive entry, in which certain privileges are negotiated at first and then are subsequently expanded. Ongoing communication with gatekeepers between the time that access is granted and the start-up of the study is recommended—this may
be a lengthy period if funding requests are involved. It is not only courteous to keep people informed, it may prove critical to the project’s success because circumstances (and leadership) at the site can change.
Timeframes Research designs designate when, and how often, data will be collected. In many studies, data are collected at one point in time. For example, patients might be asked on a single occasion to describe their health– promoting behaviors. Some designs, however, call for multiple contacts with participants, often to assess changes over time. Thus, in planning a study, researchers must decide on the number of data collection points needed to address the research question properly. The research design also designates when, relative to other events, data will be collected. For example, the design might call for weight measurements 4
and 8 weeks after an exercise intervention. Designs can be categorized in terms of study timeframes. The major distinction, for both qualitative and quantitative researchers, is between cross-sectional and longitudinal designs.
Cross-Sectional Designs Cross-sectional designs involve the collection of data once: the phenomena under study are captured at a single time point. Cross-sectional studies are appropriate for describing the status of phenomena or for describing relationships at a fixed point in time. For example, we might be interested in examining whether psychological symptoms in women going through menopause correlate contemporaneously with physiologic symptoms.
Example of a Cross-Sectional Study Van Hoek et al. (2019) studied the influence of demographic factors, resilience, and stress-reducing activities on the academic outcomes of undergraduate nursing students. Data were gathered at a single point in time from 554 Belgian nursing students.
Inferences about causal relationships are tricky when cross-sectional designs are used. For example, we might test the hypothesis, using cross-sectional data, that a determinant of excessive alcohol consumption is
low impulse control, as measured by a psychological test. When both alcohol consumption and impulse control are measured concurrently, however, it is difficult to know which variable influenced the other, if either. Cross-sectional data can best be used to infer time sequence under two circumstances: (1) when a cogent theoretical rationale guides the analysis or (2) when evidence or logic indicates that one variable preceded
the other. For example, in a study of the effects of low birth weight on morbidity in school-aged children, it is clear that birth weight came first. Cross-sectional studies can be designed to permit inferences about processes evolving over time, but such designs are weaker than longitudinal ones. Suppose, for example, we were studying changes in
children’s health promotion activities between the ages of 10 and 13 years. One way to study this would be to interview children at the age of 10 years and then 3 years later at the age of 13 years—a longitudinal design. On the other hand, we could use a cross-sectional design by interviewing dif erent children of ages 10 and 13 years and then comparing their responses. If 13-year-olds engaged in more health-promoting
activities than 10-year-olds, we might infer that children improve in making healthy choices as they age. To make this kind of inference, we would have to assume that the older children would have responded
like the younger ones had they been questioned 3 years earlier, or, conversely, that 10-year-olds would report more health-promoting activities if they were questioned again 3 years later. Such a design, which
involves a comparison of multiple age cohorts, is sometimes called a cohort comparison design. Cross-sectional studies are economical but inferring changes over time with such designs is problematic. In our example, 10- and 13-year old children may have different attitudes toward health promotion,
independent of maturation. Rapid social and technologic changes make it risky to assume that differences in the behaviors or traits of different age groups are the result of time passing rather than of cohort differences. In cross-sectional studies designed to explore change, there are often alternative explanations for the findings—and that is precisely what good research design tries to avoid.
Example of a Cross-Sectional Study With Inference of Change Over Time Hladek et al. (2018) studied the feasibility of using sweat to measure cytokines in older adults (aged 65+) compared with those in younger adults (aged 18-40 years). Higher concentrations of TNF-α and IL-10 were observed in older adults, consistent with the hypothesis that cytokines increase with age.
Longitudinal Designs A study in which researchers collect data at more than one point in time over an extended period is a longitudinal design. There are four situations in which a longitudinal design is appropriate:
1. Studying time-related processes. Some research questions specifically concern phenomena that evolve over time (e.g., wound healing). 2. Determining time sequences. It is sometimes important to establish how phenomena are sequenced. For example, if it is hypothesized that infertility affects depression, then it would be important to ascertain that the depression did not precede the fertility problem. 3. Assessing changes over time. Some studies examine whether changes have occurred over time. For example, an intervention study might examine both short-term and long-term changes in health outcomes. A qualitative study might explore the evolution of grieving in the
spouses of palliative care patients. 4. Enhancing research control. Quantitative researchers sometimes collect data at multiple points to enhance the interpretability of the results. For example, when two groups are being compared with regard to the effects of alternative interventions, the collection of preintervention data allows the researcher to assess group comparability initially.
There are several types of longitudinal designs. Most involve collecting data from one group of participants multiple times, but others involve different samples. Trend studies, for example, are investigations of a
specific phenomenon using different samples from the same population over time (e.g., every 2 years). Trend studies permit researchers to examine patterns and rates of change and to predict future developments. Many trend studies document trends in public health issues, such as smoking, obesity, and so on.
Example of a Trend Study Neaigus et al. (2017) studied trends in HIV and hepatitis C virus risk behaviors among people who inject drugs in New York City. The team examined changes from 2005 to 2009 and to 2012. Significant
trends in risk behaviors included a decline in unsafe syringe source, but an increase in vaginal or anal sex without condoms.
In a more typical longitudinal study, the same people provide data at two or more points in time. Longitudinal studies of general (nonclinical) populations are sometimes called panel studies. The term panel refers
to the sample of people providing data. Because the same people are studied over time, researchers can examine diverse patterns of change (e.g., those whose health improved or deteriorated). Panel studies are
intuitively appealing as an approach to studying change, but they are expensive.
Example of a Panel Study Many national governments sponsor large-scale panel studies whose data have been analyzed by nurse researchers. For example, Davis et al.(2018) used data from the Australian Longitudinal Study on Women’s Health to examine the relationship between parity and long-term weight gain over a 16-year period.
Follow-up studies are undertaken to examine the subsequent development of individuals who have a specified condition or who have received a specific treatment. For example, patients who have received a
special nursing intervention may be followed to ascertain long-term effects. Or, in a qualitative study, patients interviewed shortly after a diagnosis of prostate cancer may be followed to assess their experiences after treatment decisions have been made.
Example of a Qualitative Follow-Up Study Hansen et al. (2017) followed-up, over a 6-month period, the family members caring for patients with terminal hepatocellular carcinoma as patients approached the end of life. The caregivers were
interviewed monthly.
In some longitudinal studies, called cohort studies, a group of people (the cohort) is tracked over time to see if subsets with exposure to different factors diverge in terms of subsequent outcomes. For example, in
a cohort of women, those with or without a history of childbearing could be tracked to examine differences in rates of ovarian cancer. This type of study, sometimes called a prospective study, is discussed in Chapter 9. Longitudinal studies are appropriate for studying the trajectory of a phenomenon over time, but a major problem is attrition—the loss of participants after initial data collection. Attrition is problematic because
those who drop out of the study often differ in systematic ways from those who continue to participate, resulting in potential biases and difficulty in generalizing to the original population. The longer the interval between data collection points, the greater the risk of attrition and resulting biases.
In longitudinal studies, researchers make decisions about the number of data collection points and the intervals between them. When change or development is rapid, numerous time points at short intervals may
be needed to document it. Researchers interested in outcomes that may occur years after the original data collection must use longer-term follow-up—or use surrogate outcomes. For example, in evaluating the
effectiveness of a smoking cessation intervention, the main outcome of interest might by lung cancer incidence or age at death, but the researcher would likely use subsequent smoking (e.g., 3 months after the
intervention) as the surrogate outcome.
Repeated Measures Designs
Studies with multiple points of data collection are sometimes described as having a repeated measures design, which usually signifies a study in which data are collected three or more times. Longitudinal studies, such as follow-up and cohort studies, sometimes use a repeated measures design. Repeated measures designs, however, can also be used in studies that are essentially cross-sectional. For example, a study involving the collection of postoperative patient data on vital signs hourly over a 6-hour period would not be described as longitudinal because the study does not involve an extended time perspective. Yet, the design could be characterized as repeated measures. Researchers are especially likely to use the term repeated measures design when they use a repeated measures approach to statistical analysis (see Chapter 18).
Example of a Repeated Measures Design Krause-Parello et al. (2018) studied the effects of an animal-assisted intervention on hospitalized veterans receiving palliative care. Blood pressure, heart rate, and salivary cortisol were measured before,
immediately after, and again 30 minutes after the intervention.
TIP In making design decisions, you will need to balance various considerations, such as time, cost, ethics, and study rigor. Try to understand your “upper limits” before finalizing your design. That is, what
is the most money that can be spent on the project? What is the maximum amount of time available for conducting the study? What is the limit of acceptability with regard to attrition? These limits often
eliminate some design options. With these constraints in mind, the central focus should be on designing a study that maximizes the rigor or trustworthiness of the study.
Planning Data Collection
In planning a study, researchers must select methods to gather their research data. This section provides an overview of various methods of data collection for qualitative and quantitative studies. Overview of Data Collection and Data Sources A broad array of data collection methods can be used in research. In some cases, researchers may be able to use data from existing sources, such as records. Most often, however, researchers collect new data, and
one key planning decision concerns the types of data to gather. Three approaches have been used most frequently by nurse researchers: self-reports, observation, and biophysiologic measures.
Self-Reports (Patient-Reported Outcomes) A good deal of information can be gathered by questioning people directly, a method known as self-report. In the medical literature, self-reports are often called patient-reported outcomes or PROs, but some
self-reports are not about patients (e.g., self-reports about nurses’ burnout) and some are not outcomes (self-reports about prior hospitalizations). Most nursing studies involve self-report data. The unique ability of humans to communicate verbally makes direct questioning a particularly important part of nurse researchers’ data collection repertoire. Self-reports are versatile. If we want to know what people think, believe, or plan to do, the most efficient approach is to ask them. Self-reports can yield information that would be impossible to gather by other means. Behaviors can be observed but only if participants engage in them publicly. Furthermore, observers can observe only those behaviors occurring at the time of the study. Through self-reports, researchers can gather retrospective data about events occurring in the past or information about behaviors in which people plan to engage in the future. Self-reports can also capture psychological attributes such as motivation
or resilience. Nevertheless, verbal report methods have some weaknesses. The most serious issue concerns their validity and accuracy: Can we be sure that people feel or act the way they say they do? We all have a tendency
to present ourselves positively, and this may conflict with the truth. Researchers who gather self-report data should recognize this limitation and take it into consideration when interpreting the results.
Example of a Study Using Self-Reports Beattie et al. (2019) explored the perceptions of healthcare providers on workplace violence perpetrated by clients. The data came from in-depth group and one-on-one interviews with nurses and other healthcare staff in Australia.
Self-report methods depend on respondents’ willingness to share personal information. Projective techniques are sometimes used to obtain data about people’s psychological states indirectly. Projective
techniques present participants with a stimulus of low structure, permitting them to “read in” and describe their interpretations. The Rorschach (inkblot) test is an example of a projective technique. Other projective methods encourage self-expression through the construction of a product (e.g., drawings). The assumption is that people express their needs, motives, and emotions by working with or manipulating materials. Projective methods are used by nurse researchers mainly in studies exploring sensitive topics with children.
Example of a Study Using Projective Methods Anderson and Tulloch-Reid (2019) investigated the experiences of adolescents with diabetes living in Jamaica. Participants took part in group interviews and were also asked to draw pictures representing
their experiences.
Observation An alternative to self-reports is observation of study participants. Observation can be done directly through the human senses or with technical apparatus, such as video equipment, X-rays, and so on. Observational methods can be used to gather information about a wide range of phenomena, such as: (1) people’s characteristics and conditions (e.g., patients’ sleep–wake state); (2) verbal communication (e.g., nurse–patient dialogue); (3) nonverbal communication (e.g., facial expressions); (4) activities and behavior (e.g., geriatric patients’ self-grooming); (5) skill attainment (e.g., diabetic patients’ skill in testing their urine); and (6) environmental conditions (e.g., architectural barriers in nursing homes). Observation in healthcare settings is an important data-gathering strategy. Nurses are in an advantageous position to observe, relatively unobtrusively, the behaviors of patients, their families, and hospital staff. Moreover, nurses may, by training, be especially sensitive observers. Observational methods are especially useful when people are unaware of their own behavior (e.g., manifesting preoperative symptoms of anxiety), when people are embarrassed to report activities (e.g., aggressive actions), when behaviors are emotionally laden (e.g., grieving), or when people cannot describe their actions (e.g., young children). A shortcoming of observation is potential behavior distortions when
participants are aware of being observed—a problem called reactivity. Reactivity can be eliminated if observations are made without people’s knowledge, through concealment—but this may pose ethical concerns. Another problem is observer biases. Several factors (e.g., prejudices, emotions, fatigue) can undermine objectivity. Observational biases can be minimized through careful training.
Example of a Study Using Observation Vittner et al. (2018) studied whether skin-to-skin contact between parents and stable preterm infants alleviates parental stress while also supporting mother–father–infant relationships. Parent–infant
interactions were examined via video-recorded observations, in which levels of synchrony and responsiveness were recorded.
Biophysiologic Measures/Biomarkers Many clinical studies rely on the use of biophysiologic measures or biomarkers. Biomarkers are objective, quantifiable characteristics of biological processes (Strimbu & Tavel, 2010). Biophysiologic and physical variables typically are measured using specialized technical instruments and equipment. Because such equipment is available in healthcare settings, the costs of these measures to nurse researchers may be small or nonexistent. A major strength of biophysiologic measures is their objectivity. Nurse A and nurse B, reading from the same spirometer output, are likely to record the same forced expiratory volume (FEV) measurements. Furthermore, two different spirometers are likely to produce the same FEV readouts. Another advantage of physiologic measurements is the relative precision they normally offer. By relative, we are implicitly
comparing physiologic instruments with measures of psychological phenomena, such as self-report measures of anxiety or pain. Biophysiologic measures usually yield data of exceptionally high quality.
Example of a Study Using Biomarkers
Imes et al. (2019) studied factors associated with endothelial function in older adults with obstructive sleep apnea and cardiovascular disease. The variables examined included body mass index, blood
pressure, and several cholesterol values.
Records Most researchers create original data for their studies, but sometimes they take advantage of information available in records. Electronic health records and other records constitute rich data sources to which
nurse researchers may have access. Research data obtained from records are advantageous because they are economical: the collection of original data can be time-consuming and costly. Also, records avoid
problems stemming from people’s reaction to study participation. On the other hand, when researchers are not responsible for collecting data, they may be unaware of the records’ limitations and biases, such as the biases of selective deposit and selective survival. If the available
records are not the entire set of all possible such records, researchers must question how representative existing records are. Many record keepers intend to maintain an entire universe of records but may not succeed. Careful researchers should attempt to learn what biases might exist. Gregory and Radovinsky (2012) have suggested some strategies for enhancing the reliability of data extracted from medical records, and Dziadkowiec et al. (2016) have described a method of “cleaning” data extracted from electronic health records. Other difficulties also may be relevant. Sometimes records have to be verified for their authenticity or accuracy, which may be difficult if the records are old. In using records to study trends, researchers should be alert to possible changes in record-keeping procedures. Another problem is the increasing difficulty of gaining access to institutional records. Thus, although records may be plentiful and inexpensive, they should
not be used without paying attention to potential problems.
TIP Nurse researchers are increasingly using information from”Big Data” sources, such as large administrative databases or registries. Registries are collections of large amounts of data about a particular disease or patient population, such as trauma or cancer registries. Talbert and Sole (2013) and Gephart et al. (2018) have written about doing research with large databases.
Example of a Study Using Records Pressler et al. (2018) studied the symptoms, nutrition, and pressure ulcer status among older women with heart failure in relation to their return to the community from a skilled nursing facility. The data were collected from the electronic medical records.
Dimensions of Data Collection Approaches Data collection methods vary along three key dimensions: structure, researcher obtrusiveness, and objectivity. In planning a study, researchers make decisions about where on these dimensions the data collection methods should fall.
In structured data collection, information is gathered from participants in a comparable, prespecified way. Most self-administered questionnaires are structured: They include a fixed set of questions, usually with predesignated response options (e.g., agree/disagree). Structured methods give participants limited opportunities to qualify their answers or to explain the meaning of their responses. By contrast, qualitative
studies rely mainly on unstructured methods of data collection. Structured methods often take considerable effort to develop, but they yield data that are relatively easy to analyze because the data can be readily quantified. Structured methods are not appropriate for an in– depth examination of a phenomenon, however. Consider the following two methods of asking people about their levels of stress:
Structured During the past week, would you say you felt stressed:
1. rarely or none of the time, 2. some or a little of the time, 3. occasionally or a moderate amount of the time, or 4. most or all of the time?
Unstructured How stressed or anxious have you been this past week? Please tell me about any tensions and stresses you experienced. The structured question allows us to compute what percentage of respondents felt stressed most of the time but provides no information about the circumstances of the stress. The unstructured question allows
for deeper and more thoughtful responses but may not be useful for people who are not good at expressing themselves; moreover, the resulting data are more difficult to analyze.
Researcher Obtrusiveness Data collection methods differ in the degree to which people are aware of the data-gathering process. If people know they are under scrutiny, their behavior and responses may not be “normal,” and distortions can undermine the value of the research. When data are collected unobtrusively, however, ethical problems may emerge. Study participants are most likely to distort their behavior and their responses to questions under certain circumstances. Researcher obtrusiveness is likely to be most problematic when (1) a program is being
evaluated and participants have a vested interest in the evaluation outcome; (2) participants engage in socially unacceptable or unusual behavior; (3) participants have not complied with medical and nursing
instructions; and (4) participants are the type of people who have a strong need to “look good.” When researcher obtrusiveness is unavoidable under these circumstances, researchers should make an effort to put participants at ease, to emphasize the importance of candor, and to adopt a nonjudgmental demeanor. Objectivity Objectivity refers to the degree to which two independent researchers can arrive at similar “scores” or make similar observations regarding concepts of interest. Objectivity is a mechanism for avoiding biases. Some data collection approaches require more subjective judgment than others. Researchers with a positivist orientation usually strive for a reasonable amount of objectivity. In research based on the
constructivist paradigm, however, the subjective judgment of investigators is considered essential for understanding human experiences. Developing a Data Collection Plan
In planning a study, researchers make decisions about the type and amount of data to collect. Several factors, including costs, must be weighed, but a key goal is to identify the kinds of data that will yield
accurate, valid, and trustworthy information for addressing the research question. Most researchers face the issue of balancing information needs against the risk of overburdening participants. In many studies, more data are collected than are needed or analyzed. Although it is better to have adequate data than to have unwanted omissions, minimizing participant burden should be an important goal. Specific guidance on data collection plans is offered in Chapter 14 for quantitative studies and Chapter 24 for qualitative studies. Organization of a Research Project
Studies typically take many months to complete and longitudinal studies require years of work. During the planning phase, it is a good idea to make preliminary estimates of how long various tasks will require. Having deadlines helps to restrict tasks that might otherwise continue indefinitely, such as a literature review. Chapter 3 presented a sequence of steps that quantitative researchers follow in a study. The steps represented an idealized conception: the research process rarely follows a neatly prescribed sequence of procedures, even in quantitative studies. Decisions made in one step, for example, may require alterations in a previous activity. For example, sample size decisions may require rethinking how many sites are needed. Nevertheless, preliminary time estimates are valuable. In particular, it is important to have a sense of how much total time the study will require and when it will begin.
TIP We could not suggest even approximations for the percentage of time that should be spent on each task. Some projects need many months to recruit participants, whereas other studies can rely on an
existing group. Clearly, not all steps are equally time-consuming.
Researchers sometimes develop visual timelines to help them organize a study. These devices are especially useful if funding is sought because the schedule helps researchers to understand when and for how
long staff support is needed (e.g., for transcribing interviews). This can best be illustrated with an example, in this case of a hypothetical quantitative study. Suppose a researcher was studying the following problem: Is a woman’s decision to have an annual mammogram related to her perceived susceptibility to breast cancer? Using the organization of steps outlined
in Chapter 3, here are some of the tasks that might be undertaken: a
1. The researcher is concerned that many older women do not get mammograms regularly. Her specific research question is whether mammogram practices are different for women with different perceptions about their susceptibility to breast cancer. 2. The researcher reviews the research literature on breast cancer, mammography use, and factors affecting mammography decisions. 3. The researcher does clinical fieldwork by discussing the problem with nurses and other healthcare professionals in various clinical settings and by having informal discussions with women in a support group for breast cancer patients.
4. The researcher seeks theories and models for her problem. She finds that the Health Belief Model is relevant, which helps her to develop a conceptual definition of susceptibility to breast cancer. 5. Based on the framework, the following hypothesis is developed: Women (P) who perceive themselves as susceptible to breast cancer (I) are more likely than other women (C) to get an annual mammogram (O). 6. The researcher adopts a nonexperimental, cross-sectional, between-subjects research design. Her comparison strategy will be to compare women with different rankings on a measure of susceptibility to breast cancer. She designs the study to control the confounding variables of age, marital status, and health insurance status. Her research site will be Pittsburgh. 7. There is no intervention in this study and so this step is unnecessary. 8. The researcher designates that the population of interest is women between the ages of 50 and 65 years living in Pittsburgh who have not been previously diagnosed as having any form of cancer. 9. The researcher will recruit 250 women living in Pittsburgh as her research sample; they are identified at random using a procedure known as random-digit dialing, and so she does not need to gain entrée into any institution. 10. Research variables will be measured by self-report; the independent variable (perceived susceptibility), dependent variable (mammogram history), and confounding variables will be measured by asking participants a series of questions. 11. The Institutional Review Board (IRB) at the researcher’s institution is asked to review the plans to ensure that the study adheres to ethical standards. 12. Plans for the study are finalized: the methods are reviewed by colleagues with clinical and methodologic expertise and by the IRB; the data collection instruments are pretested; and interviewers who will collect the data are trained. 13. Data are collected by means of telephone interviews with women in the research sample. 14. Data are prepared for analysis by coding them and entering them onto a computer file. 15. Data are analyzed using statistical software. 16. The results indicate that the hypothesis is supported; however, the researcher’s interpretation must take into consideration that many women who were asked to participate declined to do so. 17. The researcher presents an early report on her findings and interpretations at a conference of Sigma Theta Tau International. She subsequently publishes the report in the International Journal of Nursing Studies. 18. The researcher seeks out clinicians to discuss how the study findings can be used in practice.
The researcher plans to conduct this study over a 2-year period; Figure 8.2 presents a hypothetical schedule. Many steps overlap or are undertaken concurrently; some steps are projected to involve little time, whereas others require months of work. (The Toolkit in the accompanying Resource Manual includes Figure 8.2 as a Word document for you to adapt.)
FIGURE 8.2 Project timeline (in months) for a hypothetical study of women’s mammography decisions.
In developing a schedule, several considerations should be kept in mind, including methodologic expertise and the availability of funding. In the present example, if the researcher needed financial support to pay
for the cost of interviewers, the timeline would need to be expanded to accommodate the time required to prepare a proposal and await the funding decision. It is also important to consider the practical aspects of performing the study, which were not noted in the preceding section. Securing permissions, hiring staff, and holding meetings are all time-consuming, but necessary, activities.
In large-scale studies—especially studies in which there is an intervention—it is wise to undertake a pilot study. A pilot study is a trial run designed to test planned methods and procedures. Results and
experiences from pilot studies help to inform many of decisions for larger projects. We discuss the important role of pilot studies in Chapter 29.
Individuals differ in the kinds of tasks that appeal to them. Some people enjoy the preliminary phase, which has an intellectual component; others are more eager to collect the data, which is more interpersonal. Researchers should, however, allocate a sensible amount of time to do justice to each activity.
TIP Getting organized for a study has many dimensions beyond having a timeline. Two important issues concern having the right team and mix of skills for a research project, and developing plans for hiring and monitoring research staff (Nelson & Morrison-Beedy, 2008).
Critical Appraisal of the Planning Aspects of a Study
Researchers typically do not describe the planning process or problems that arose during the study in journal articles. Thus, there is typically little that readers can do to critically appraise the researcher’s planning efforts. What can be appraised, of course, are the outcomes of the planning—that is, the methodologic decisions themselves. Guidelines for critically appraising those decisions are provided throughout
this book. Readers can, however, be alert to a few things relating to research planning. First, evidence of careful conceptualization provides a clue that the project was well planned. If a conceptual map is presented (or
implied) in the report, it means that the researcher had a “road map” that facilitated planning. Second, readers can consider whether the researcher’s plans reflect adequate attention to concerns about evidence-based practice. For example, was the comparison group strategy designed to reflect a realistic practice concern? Was the setting one that maximizes potential for the generalizability of the findings? Did the timing of data collection correspond to clinically important milestones? Was the intervention
sensitive to the constraints of a typical practice environment? Finally, a report might provide clues about whether the researcher devoted sufficient time and resources in preparing for the study. For example, if the report indicates that the study grew out of earlier research
on a similar topic, or that the researcher had previously used the same instruments, or had completed other studies in the same setting, this suggests that the researcher was not plunging into unfamiliar waters. Unrealistic planning can sometimes be inferred from a discussion of sample recruitment. If the report indicates that the researcher was unable to recruit the originally hoped-for number of participants, or if
recruitment took months longer than anticipated, this suggests that the researcher may not have done adequate homework during the planning phase.
Research Example
In this section, we describe a pilot study and the “lessons learned” by the researchers. This is a good example of the importance of strong advance planning for a study.
Study: Recruitment of older African American males for depression research: Lessons learned (Bryant et al., 2014) Purpose: The purpose of the article was to describe the setbacks and lessons learned in a pilot study aimed at exploring the signs and symptoms of depression experienced by older African American men. Methods: The researchers sought to recruit a sample of about 20 African American men aged 60 years and older over a 3 to 4-month recruitment period. The men were to have been interviewed to learn how they
recognize, express, and describe their depression. Initial recruitment was through flyers distributed to community clinics and physicians’ offices serving the target group. The colorful flyers included photos and a description of the study and contact information. Findings: Nine months into recruitment, only one person had inquired about participation in the study, and that person was deemed ineligible. This recruitment failure prompted members of the team to solicit
feedback from university community liaisons and a local community development group. The advisers thought the study was important, but noted that the researchers faced numerous recruitment barriers, such
as the likelihood that older black men would not easily trust outsiders and might believe that they are too strong to be depressed. The advisers also provided valuable feedback about the recruitment flier and
other aspects of the study design. Conclusions: The researchers concluded that their “failure to recruit participants can be ascribed to a number of missteps: non-culturally relevant recruitment materials, a failure to build trust and engage
community coalitions beforehand, (and) the use of ineffective strategies to address the stigma associated with mental illness” (p. 4). They noted that the lessons learned would hopefully facilitate future
recruitment efforts for mental health research involving black men.
Summary Points
Researchers face numerous challenges in planning a study, including the challenge of designing a study that is strong with respect to reliability and validity (quantitative studies) or trustworthiness (qualitative studies). Reliability refers to the accuracy and consistency of information obtained in a study. Validity is a more complex concept that broadly concerns the soundness of the study’s evidence—that is, whether the findings are cogent and well grounded. Trustworthiness in qualitative research encompasses several different dimensions, including dependability, confirmability, authenticity, transferability, and credibility. Credibility is achieved to the extent that the research methods engender confidence in the truth of the data and in the researchers’ interpretations. Triangulation, the use of multiple sources or referents to draw conclusions about what constitutes the truth, is one approach to
enhancing credibility. A bias is an influence that distorts study results. Systematic bias results when a bias operates in a consistent direction.
In quantitative studies, research control is used to hold constant outside influences on the outcome variable so that its relationship to the independent variable can be better understood. Researchers use various strategies to control confounding variables, which are extraneous
to the study aims and can obscure understanding.
In quantitative studies, a powerful tool to eliminate bias is randomness—having certain features of the study established by chance rather than by researchers’ intentions. Reflexivity, the process of reflecting critically on the self and of scrutinizing personal values that could affect interpretation, is an important tool in qualitative research. Generalizability in a quantitative study concerns the extent to which findings can be applied to people or settings other than the ones used in the research. Transferability is the extent to which qualitative findings can be transferred to other settings. During the planning phase, researchers need to consider the extent to which key stakeholders will be involved in the research and who the key stakeholders are.
In planning a study, researchers make many design decisions, including whether to have an intervention, how to control confounding variables, what type of comparisons will be made, where the study will take place, and what the study timeframes will be. Quantitative researchers often incorporate comparisons into their designs to enhance interpretability. In between-subjects designs, different groups of people are compared. Within-subjects designs involve comparisons of the same people at different times or under different
circumstances, and mixed designs involve both types of comparison. Site selection for a study often requires site visits to evaluate suitability and feasibility. Gaining entrée into a site involves developing and maintaining trust with gatekeepers. Cross-sectional designs involve collecting data at one point in time, whereas longitudinal designs involve data collection two or more times over an extended period. Trend studies have multiple points of data collection with different samples from the same population. Panel studies gather data from the same people, usually from a general population, more than once. In a follow-up study, data are gathered two or more times from a well-defined group (e.g., those with a particular health problem). In a cohort study, a cohort of people is tracked over time to see if subsets with different exposures to risk factors differ in terms of subsequent outcomes. A repeated measures design typically involves collecting data three or more times, either in a longitudinal fashion or in rapid succession over a shorter timeframe. Longitudinal studies are typically expensive and time-consuming, and have risk of attrition (loss of participants over time) but are essential for illuminating time-related phenomena. Researchers also develop a data collection plan. In nursing, the most widely used methods are self-report, observation, biophysiological measures, and existing records. Self-report data (sometimes called patient-reported outcomes or PROs) are obtained by directly questioning people. Self-reports are versatile and powerful but a drawback is the potential for respondents’ deliberate or inadvertent misrepresentations. A wide variety of human activity and traits are amenable to direct observation. Observation is subject to observer biases and distorted participant behavior (reactivity). Biophysiologic measures (biomarkers) tend to yield high-quality data that are objective and valid. Existing records and documents are an economical source of research data, but two potential biases in records are selective deposit and selective survival. Data collection methods vary in terms of structure, researcher obtrusiveness, and objectivity, and researchers must decide on these dimensions in their plan. Planning efforts should include the development of a timeline that provides estimates of when important tasks will be completed.
Study Activities
Study activities are available to instructors on .
References Cited in Chapter 8
Anderson M., & Tulloch-Reid M. (2019). “You cannot cure it, just control it”: Jamaican adolescents living with diabetes. Comprehensive Child and Adolescent Nursing, 42(2), 109–123. Beattie J., Griffiths D., Innes K., & Morphet J. (2019). Workplace violence perpetrated by clients of health care: A need for safety and trauma-informed care. Journal of Clinical Nursing, 28, 116–124. Bower R., Coad J., Manning J., & Pengelly T. (2018). A qualitative, exploratory study of nurses’ decision-making when interrupted during medication administration within the paediatric intensive care unit. Intensive & Critical Care Nursing, 44, 11–17.
* Bryant K., Wicks M., & Willis N. (2014). Recruitment of older African American males for depression research: Lessons learned. Archives of Psychiatric Nursing, 28, 17–20.
* Buckwalter K., Grey M., Bowers B., McCarthy A., Gross D., Funk M., & Beck C. (2009). Intervention research in highly unstable environments. Research in Nursing & Health, 32, 110–121.
* Concannon T., Meissner P., Grunbaum J., McElwee N., Guise J. M., Santa J., … Leslie L. (2012). A new taxonomy for stakeholder engagement in patient-centered outcomes research. Journal of General Internal Medicine, 27, 985–991. Currie G., & Szabo J. (2019). “It is like a jungle gym, and everything is under consideration”: The parent’s perspective of caring for a child with a rare disease. Child: Care, Health and Development, 45, 96–103. Davis D., Brown W., Foureur M., Nohr E., & Xu F. (2018). Long-term weight gain and risk of overweight in parous and nulliparous women. Obesity, 26, 1072–1077. Dias N., Brandon D., Haase J., & Tanabe P. (2018). Bereaved parents’ health status during the first 6 months after their child’s death. American Journal of Hospice & Palliative Care, 35, 829–839.
* Dziadkowiec O., Callahan T., Ozkaynak M., Reeder B., & Welton J. (2016). Using a data quality framework to clean data extracted from the electronic health record: A case study. EGEMS, 4, 1201. Gephart S., Davis M., & Shea K. (2018). Perspectives on policy and the value of nursing science in a Big Data era. Nursing Science Quarterly, 31, 78–81.
* Gregory K. E., & Radovinsky L. (2012). Research strategies that result in optimal data collection from the patient medical record. Applied Nursing Research, 25, 108–116. Hansen L., Rosenkranz S., Wherity K., & Sasaki A. (2017). Living with hepatocellular carcinoma near the end of life: Family caregivers’ perspectives. Oncology Nursing Forum, 44, 562–570. Hladek M., Szanton S., Cho Y., Lai C., Sacko C., Roberts L., & Gill J. (2018). Using sweat to measure cytokines in older adults compared to younger adults. Journal of Immunological Methods, 454, 1–5.
** Imes C., Baniak L., Choi J., Luyster F., Morris J., Ren D., & Chasens E. (2019). Correlates of endothelial function in older adults with untreated obstructive sleep apnea and cardiovascular disease. Journal of Cardiovascular Nursing, 34, E1–E7. Krause-Parello C., Levy C., Holman E., & Kolassa J. (2018). Effects of VA facility dog on hospitalized veterans seen by a palliative care psychologist. American Journal of Hospice & Palliative Care, 35, 5–14. Lincoln Y. S., & Guba E. G. (1985). Naturalistic inquiry. Newbury Park, CA: Sage. Neaigus A., Reilly K., Jenness S., Hagan H., Wendel T., Gelpi-Acosta C., & Marshall D. (2017). Trends in HIV and HVC risk behaviors and prevalent infection among people who inject drugs in New York City, 2005-2012. Journal of Acquired Immune Deficiency Syndromes, 75, S325–S332.
* Nelson L. E., & Morrison-Beedy D. (2008). Research team training: moving beyond job descriptions. Applied Nursing Research, 21, 159–164. Patton M. Q. (2015). Qualitative research & evaluation methods (4th ed.). Thousand Oaks, CA: Sage. Polit D. F., & Beck C. T. (2010). Generalization in qualitative and quantitative research: Myths and strategies. International Journal of Nursing Studies, 47, 1451–1458. Pressler S., Jung M., Titler M., Harrison J., & Lee K. (2018). Symptoms, nutrition, pressure ulcers, and return to community among older women with heart failure at skilled nursing facilities. Journal of Cardiovascular Nursing, 33, 22–29. Schwandt T. (2007). The Sage dictionary of qualitative inquiry (3rd ed.). Thousand Oaks, CA: Sage.
* Strimbu K., & Tavel J. (2010). What are biomarkers? Current Opinion in HIV & AIDS, 5, 463–366. Talbert S., & Sole M. L. (2013). Too much information: Research issues associated with large databases. Clinical Nurse Specialist, 27, 73–80. Van Hoek G., Portzky M., & Franck E. (2019). The influence of sociodemographic factors, resilience and stress-reducing activities on academic outcomes of undergraduate nursing students: A cross-sectional research study. Nurse Education Today, 72, 90–96. Van der Meulen I., May A., Koole R., & Ros W. (2018). A distress thermometer intervention for patients with head and neck cancer. Oncology Nursing Forum, 45, E14–E32. Vittner D., McGrath J., Robinson J., Lawhon G., Cusson R., Eisenfeld L., … Cong X. (2018). Increase in oxytocin from skin-to-skin contact enhances development of parent-infant relationship. Biological Research for Nursing, 20, 54–62.
*A link to this open-access article is provided in the Toolkit for Chapter 8 in the Resource Manual.
**This journal article is available on for this chapter.
aThis is only a partial list of tasks and is designed to illustrate the flow of activities; the flow in this example is more orderly than would ordinarily be true.
Designing and Conducting Quantitative Studies to Generate Evidence for Nursing
Chapter 9 Quantitative Research Design Chapter 10 Rigor and Validity in Quantitative Research Chapter 11 Specific Types of Quantitative Research Chapter 12 Quality Improvement and Improvement Science Chapter 13 Sampling in Quantitative Research Chapter 14 Data Collection in Quantitative Research Chapter 15 Measurement and Data Quality Chapter 16 Developing and Testing Self-Report Scales Chapter 17 Descriptive Statistics Chapter 18 Inferential Statistics Chapter 19 Multivariate Statistics Chapter 20 Processes of Quantitative Data Analysis Chapter 21 Clinical Significance and Interpretation of Quantitative Results

Get Professional Assignment Help Cheaply

Buy Custom Essay

Don't use plagiarized sources. Get Your Custom Essay on
Research Problems, Research Questions, and Hypotheses
Just from $10/Page
Order Essay

Are you busy and do not have time to handle your assignment? Are you scared that your paper will not make the grade? Do you have responsibilities that may hinder you from turning in your assignment on time? Are you tired and can barely handle your assignment? Are your grades inconsistent?

Whichever your reason is, it is valid! You can get professional academic help from our service at affordable rates. We have a team of professional academic writers who can handle all your assignments.

Why Choose Our Academic Writing Service?

  • Plagiarism free papers
  • Timely delivery
  • Any deadline
  • Skilled, Experienced Native English Writers
  • Subject-relevant academic writer
  • Adherence to paper instructions
  • Ability to tackle bulk assignments
  • Reasonable prices
  • 24/7 Customer Support
  • Get superb grades consistently

Online Academic Help With Different Subjects


Students barely have time to read. We got you! Have your literature essay or book review written without having the hassle of reading the book. You can get your literature paper custom-written for you by our literature specialists.


Do you struggle with finance? No need to torture yourself if finance is not your cup of tea. You can order your finance paper from our academic writing service and get 100% original work from competent finance experts.

Computer science

Computer science is a tough subject. Fortunately, our computer science experts are up to the match. No need to stress and have sleepless nights. Our academic writers will tackle all your computer science assignments and deliver them on time. Let us handle all your python, java, ruby, JavaScript, php , C+ assignments!


While psychology may be an interesting subject, you may lack sufficient time to handle your assignments. Don’t despair; by using our academic writing service, you can be assured of perfect grades. Moreover, your grades will be consistent.


Engineering is quite a demanding subject. Students face a lot of pressure and barely have enough time to do what they love to do. Our academic writing service got you covered! Our engineering specialists follow the paper instructions and ensure timely delivery of the paper.


In the nursing course, you may have difficulties with literature reviews, annotated bibliographies, critical essays, and other assignments. Our nursing assignment writers will offer you professional nursing paper help at low prices.


Truth be told, sociology papers can be quite exhausting. Our academic writing service relieves you of fatigue, pressure, and stress. You can relax and have peace of mind as our academic writers handle your sociology assignment.


We take pride in having some of the best business writers in the industry. Our business writers have a lot of experience in the field. They are reliable, and you can be assured of a high-grade paper. They are able to handle business papers of any subject, length, deadline, and difficulty!


We boast of having some of the most experienced statistics experts in the industry. Our statistics experts have diverse skills, expertise, and knowledge to handle any kind of assignment. They have access to all kinds of software to get your assignment done.


Writing a law essay may prove to be an insurmountable obstacle, especially when you need to know the peculiarities of the legislative framework. Take advantage of our top-notch law specialists and get superb grades and 100% satisfaction.

What discipline/subjects do you deal in?

We have highlighted some of the most popular subjects we handle above. Those are just a tip of the iceberg. We deal in all academic disciplines since our writers are as diverse. They have been drawn from across all disciplines, and orders are assigned to those writers believed to be the best in the field. In a nutshell, there is no task we cannot handle; all you need to do is place your order with us. As long as your instructions are clear, just trust we shall deliver irrespective of the discipline.

Are your writers competent enough to handle my paper?

Our essay writers are graduates with bachelor's, masters, Ph.D., and doctorate degrees in various subjects. The minimum requirement to be an essay writer with our essay writing service is to have a college degree. All our academic writers have a minimum of two years of academic writing. We have a stringent recruitment process to ensure that we get only the most competent essay writers in the industry. We also ensure that the writers are handsomely compensated for their value. The majority of our writers are native English speakers. As such, the fluency of language and grammar is impeccable.

What if I don’t like the paper?

There is a very low likelihood that you won’t like the paper.

Reasons being:

  • When assigning your order, we match the paper’s discipline with the writer’s field/specialization. Since all our writers are graduates, we match the paper’s subject with the field the writer studied. For instance, if it’s a nursing paper, only a nursing graduate and writer will handle it. Furthermore, all our writers have academic writing experience and top-notch research skills.
  • We have a quality assurance that reviews the paper before it gets to you. As such, we ensure that you get a paper that meets the required standard and will most definitely make the grade.

In the event that you don’t like your paper:

  • The writer will revise the paper up to your pleasing. You have unlimited revisions. You simply need to highlight what specifically you don’t like about the paper, and the writer will make the amendments. The paper will be revised until you are satisfied. Revisions are free of charge
  • We will have a different writer write the paper from scratch.
  • Last resort, if the above does not work, we will refund your money.

Will the professor find out I didn’t write the paper myself?

Not at all. All papers are written from scratch. There is no way your tutor or instructor will realize that you did not write the paper yourself. In fact, we recommend using our assignment help services for consistent results.

What if the paper is plagiarized?

We check all papers for plagiarism before we submit them. We use powerful plagiarism checking software such as SafeAssign, LopesWrite, and Turnitin. We also upload the plagiarism report so that you can review it. We understand that plagiarism is academic suicide. We would not take the risk of submitting plagiarized work and jeopardize your academic journey. Furthermore, we do not sell or use prewritten papers, and each paper is written from scratch.

When will I get my paper?

You determine when you get the paper by setting the deadline when placing the order. All papers are delivered within the deadline. We are well aware that we operate in a time-sensitive industry. As such, we have laid out strategies to ensure that the client receives the paper on time and they never miss the deadline. We understand that papers that are submitted late have some points deducted. We do not want you to miss any points due to late submission. We work on beating deadlines by huge margins in order to ensure that you have ample time to review the paper before you submit it.

Will anyone find out that I used your services?

We have a privacy and confidentiality policy that guides our work. We NEVER share any customer information with third parties. Noone will ever know that you used our assignment help services. It’s only between you and us. We are bound by our policies to protect the customer’s identity and information. All your information, such as your names, phone number, email, order information, and so on, are protected. We have robust security systems that ensure that your data is protected. Hacking our systems is close to impossible, and it has never happened.

How our Assignment Help Service Works

1. Place an order

You fill all the paper instructions in the order form. Make sure you include all the helpful materials so that our academic writers can deliver the perfect paper. It will also help to eliminate unnecessary revisions.

2. Pay for the order

Proceed to pay for the paper so that it can be assigned to one of our expert academic writers. The paper subject is matched with the writer’s area of specialization.

3. Track the progress

You communicate with the writer and know about the progress of the paper. The client can ask the writer for drafts of the paper. The client can upload extra material and include additional instructions from the lecturer. Receive a paper.

4. Download the paper

The paper is sent to your email and uploaded to your personal account. You also get a plagiarism report attached to your paper.

smile and order essay GET A PERFECT SCORE!!! smile and order essay Buy Custom Essay

Place your order
(550 words)

Approximate price: $22

Calculate the price of your order

550 words
We'll send you the first draft for approval by September 11, 2018 at 10:52 AM
Total price:
The price is based on these factors:
Academic level
Number of pages
Basic features
  • Free title page and bibliography
  • Unlimited revisions
  • Plagiarism-free guarantee
  • Money-back guarantee
  • 24/7 support
On-demand options
  • Writer’s samples
  • Part-by-part delivery
  • Overnight delivery
  • Copies of used sources
  • Expert Proofreading
Paper format
  • 275 words per page
  • 12 pt Arial/Times New Roman
  • Double line spacing
  • Any citation style (APA, MLA, Chicago/Turabian, Harvard)

Our guarantees

Delivering a high-quality product at a reasonable price is not enough anymore.
That’s why we have developed 5 beneficial guarantees that will make your experience with our service enjoyable, easy, and safe.

Money-back guarantee

You have to be 100% sure of the quality of your product to give a money-back guarantee. This describes us perfectly. Make sure that this guarantee is totally transparent.

Read more

Zero-plagiarism guarantee

Each paper is composed from scratch, according to your instructions. It is then checked by our plagiarism-detection software. There is no gap where plagiarism could squeeze in.

Read more

Free-revision policy

Thanks to our free revisions, there is no way for you to be unsatisfied. We will work on your paper until you are completely happy with the result.

Read more

Privacy policy

Your email is safe, as we store it according to international data protection rules. Your bank details are secure, as we use only reliable payment systems.

Read more

Fair-cooperation guarantee

By sending us your money, you buy the service we provide. Check out our terms and conditions if you prefer business talks to be laid out in official language.

Read more
error: Content is protected !!
Open chat
Need assignment help? You can contact our live agent via WhatsApp using +1 718 717 2861

Feel free to ask questions, clarifications, or discounts available when placing an order.
  +1 718 717 2861           + 44 161 818 7126           [email protected]
 +1 718 717 2861         [email protected]