Nursing Research, Chapter 37 by Kathryn A. Lauchner, RN, Ph. D.

Imagine that you are in the library, and you take a break from studying your fundamentals text. In the study carrel you find a nursing research journal and, as you flip through the journal, you come upon the following text:

"The purpose of this study was to determine the accuracy and reliability of three types of thermometers: IVAC (IVAC Corp., San Diego, CA), TempraDOT (PyMaH Corp., Somerville, NJ), and an off-the-shelf mercury thermometer used to take oral temperatures. Roy's Adaptation Model provided the theoretical basis for the study. The convenience sample consisted of 35 adults who volunteered for the study. Information was collected by two registered nurses following the manufacturers' recommendations required for an accurate reading. Results indicated no differences in the validity and reliability of the three types of thermometers."

Just then, your instructor stops by your carrel and says, "I see you're reading the study about thermometers." She asks you the following questions:

Do you think a sample of 35 adults is large enough that you could use this research with adult clients?

If you were working in pediatrics, could you use this study to guide your practice with children?

Do other factors besides the recommendations made by the manufacturer affect the accuracy of temperature measurement?

How do you know which of the temperature readings is the correct one if the readings varied from one thermometer to another?

If you were going to take rectal or axillary temperatures, would an IVAC and a mercury thermometer be equally accurate?

How would you answer?

Welcome to the world of nursing research. It shapes your daily practice in ways you never imagined. Indeed, it forms the basis for most of the nursing knowledge you will use in your nursing career and for many of the facts and concepts presented in your nursing courses.

Research has the potential to improve nursing practice. However, before you can begin to incorporate research successfully into your practice, you must be able to evaluate research. Questions such as those asked above and many others need to be answered before you can accept any research study as a valid basis to change your practice. The purpose of this chapter is to help you ask the right questions and evaluate the answers before you use nursing research in your practice.

IMPORTANCE OF NURSING RESEARCH

Nursing research is the method used to develop or search for knowledge about issues important to nurses and nursing practice today and into the future. More than ever before, nurses are being asked to determine and document their role in the delivery of health care. To define that role, nurses need a distinct body of knowledge that separates nursing from other health professions. Nurses must be able to articulate what they do and how their actions improve the health of their clients.

Additionally, research can provide a guide when making the many decisions inherent in nursing practice. Research can help you answer such important questions as: What should I assess? Which clients can reach certain goals? Which interventions produce the best client outcomes? When should evaluation take place? Nursing research also has the potential to identify the importance of the caring component of health care delivery in achieving the outcome of a healthier society.

Historical Perspective

The first acknowledged nurse researcher was Florence Nightingale, who systematically demonstrated the importance of nursing in reducing the morbidity of wounded soldiers in the Crimean war (Rogers, 1989). When she arrived at the British military hospital at Scutari in November 1854, Nightingale found a hospital that was designed for 1,700 clients but filled with nearly 4,000 people (Fig. 18–1).

The hospital had no furniture, eating utensils, or blankets, but it did have lice, maggots, rats, and other vermin. Men lay naked on the floor in their own excrement, and filth covered the walls. It was later determined that 75% of all the casualties suffered by the British army resulted from diseases contracted in the hospital, such as dysentery, typhoid, and cholera (Kalisch & Kalisch, 1978).

Nightingale implemented reforms and collected information on the factors that influenced soldier morbidity and mortality. She considered the effects of cleanliness, ventilation, nutrition, temperature, and humidity. When she arrived, the mortality rate at the hospital was 60 percent. When she left, the mortality rate was just over 1 percent. Her interventions reduced the total mortality rate from 42 percent in February 1885 to 2 percent in June of the same year (Hebert, 1981).

Nightingale presented her findings in tables, charts, and graphs that were quite sophisticated by 1850s standards (Rogers, 1989). To influence Parliament and the British army, she wrote a study entitled Notes on Matters Affecting the Health, Efficiency and Hospital Administration of the British Army, which was later published as Notes on Hospitals (1859). As a result of Nightingale's research, the views of the British army changed.

From the time of Nightingale until the 1950s, few nurses conducted nursing research. That was largely because the nursing education did not include information about research and research techniques. In the 1920s, The American Journal of Nursing began publishing case studies, which were the earliest form of published nursing research. Case studies involved the analysis of one or more clients to increase nursing knowledge of clients with similar conditions.

During the 1950s, nursing research focused on nursing education, primarily because nurses with the knowledge to conduct nursing research were nurse educators. With the beginning of clinical specialty groups, such as medical-surgical, psychiatric, pediatric, and obstetric, nursing research began addressing specific client care issues. The combination of all of these factors promoted nursing research.

In 1965, the American Nurses' Association (ANA) sponsored the first in a series of research conferences. These conferences stimulated an increasing number of clinical studies to focus on the quality of client care and nursing interventions. These conferences continue today, along with many others at which nursing research is presented.

The 1970s were marked by a growth in nursing research. The number of nurses with master's degrees in practitioner and clinical nurse specialist programs increased rapidly, along with the number working at the doctoral level. These nurses had the knowledge to conduct research and to use research in nursing practice.

The conduct of clinical nursing research was the primary focus of the 1980s. In 1980, the American Nurses' Association Commission on Nursing Research identified priorities that helped focus research on clinical nursing practice. In 1985, the same group, known as the ANA Cabinet on Nursing Research, expanded the priorities for nursing research. Currently, the ANA Council on Nursing Research has established the priorities presented in Box 18–1.

The number of clinical journals publishing nursing research and the body of nursing knowledge generated through research have increased rapidly. In 1985, the National Center for Nursing Research (NCNR) was created under the National Institutes of Health. The purpose of the NCNR was "the conduct, support, and dissemination of information regarding basic clinical nursing research, training, and other programs in patient care research" (Bauknecht, 1985, p. 2).

During the early 1990s, the NCNR brought together prominent nurse scientists for two Conferences on Research Priorities (CORP). CORP priorities are listed in Box 18–2. In 1993, the NCNR became the National Institute of Nursing Research (NINR). The change in title increased the recognition of nursing as a research-based health profession.

Today, nursing research is focusing on health promotion and illness prevention as well as traditional health restoration. The number of clinical nursing studies continues to increase. Nurses are developing an increasingly strong scientific base of knowledge for nursing practice.

ETHICS OF NURSING RESEARCH

Ethical standards are especially important in nursing research because studies of human health patterns involve the use of human subjects whose rights must be protected. Many of these subjects are especially vulnerable. For example, if psychiatric clients are the subject of a study on medication compliance, the researcher is ethically obliged to question the ability of these clients to fully consent to the study. Infants, children, clients in pain, terminally ill clients, and very old clients are other especially vulnerable populations. The Tuskegee syphilis study illustrates the importance of maintaining ethical standards in research.

The Tuskegee study was conducted between 1938 and 1972 and was supported by the U.S. Public Health Service. It involved 400 poor African-American men from Tuskegee, Alabama, who had contracted syphilis. Their medical treatment was withheld so researchers could study the course of the disease. The subjects were examined periodically but did not receive treatment, even after penicillin was found to be effective. Many of the subjects who consented to participate in the study were not informed about the purpose and procedures of the study. Some were unaware that they were subjects at all (Vessey & Gennaro, 1994). No effort was made to stop the study, even though findings from the study were published every 4 to 6 years. As late as 1969, the Centers for Disease Control decided that the study should continue.

Too often, researchers' desires for knowledge or advancement have led to injury, increased illness, and even death among research subjects. Consequently, researchers now must adhere to strict guidelines intended to protect those who take part in studies.

Codes of Ethics

To protect research subjects from ethical violations, many disciplines have established their own codes of ethics. A code of ethics serves to guide researchers in protecting the rights of human subjects. In 1975, the ANA published Human Rights Guidelines for Nurses in Clinical and Other Research (Box 18–3). Likewise, the American Sociological Association (1984) and the American Psychological Association (1982) published codes of ethics for sociological and psychological research. Although the principles of ethical human research are common across disciplines, these three documents each address the particular concerns of their disciplines.

Risk/Benefit Ratio

Conducting ethical research involving human subjects requires analyzing the risk/benefit ratio. This ratio measures whether the risks to research subjects outweigh the possible benefits to society and the nursing profession. The risk to subjects should never exceed the value to society. For nurse researchers, this principle can be interpreted to mean that the topic of the research should be significant and have the potential to improve client care. If the research topic does not have the potential to improve client care, the research should not be conducted.

Review Boards

Most hospitals, universities, and other institutions where research is frequently conducted have formed review boards to protect the rights of subjects. These review boards are formal committees with protocols for reviewing research proposals and plans. Review boards are sometimes called human subjects committees.

Research supported by federal funds is subject to strict guidelines concerning the use of human or animal subjects. An institutional review board (IRB) is a committee whose duties include making sure that proposed research meets the federal requirements for ethical research. The committee is mandated by the federal government if the institution is receiving federal funds for research.

Informed Consent

Informed consent means that the subjects have been given sufficient information about the research to enable them to consent voluntarily to participate or decline to participate. Subjects must have the ability to comprehend the information they are given about the study. They also have the right to know the following (Polit & Hungler, 1995):

Most researchers will provide subjects with a written consent form, which the researcher will ask the subject to sign before joining the study. This document is kept by the researcher, and a copy is retained by the subject. The above procedure protects the rights of most subjects, but sometimes researchers are interested in studying special vulnerable groups of subjects who may be incapable of giving fully informed consent. Vulnerable groups typically include infants; children; mentally, emotionally, or physically disabled individuals; pregnant women; the terminally ill; and institutionalized or hospitalized clients. With these vulnerable subjects, the researcher may be required to undertake additional procedures and be especially sensitive about protecting subjects' rights. The researcher should make sure that risks are minimal and possible benefits are pronounced.

THE RESEARCH STUDY

The research project itself is usually called a study or an investigation. The tools that the researcher uses to conduct a study are called instruments. Instruments can be thermometers, laboratory tests, a list of questions, or a checklist. The term data is used to designate the information the researcher is interested in collecting. Data may take the form of words, or they may have numeric value.

The people being studied are called subjects, study participants, or informants. Subjects are enlisted from populations, the groups of individuals that the researcher is studying. For example, if a researcher wants to understand the reasons abused wives stay in relationship with their abusers, a sample of abused wives will be asked to participate in the study. If the sample truly represents the population of abused wives, the researcher can derive conclusions about the whole population from the sample.

Personnel

The person who undertakes a research project is usually called a researcher but is sometimes called an investigator. When a team of people undertakes a research project, the person directing the study is called the principal investigator or project director.

Types of Studies

The type of research depends on the questions asked by the researcher and the topic under study. Some research questions may be asked in different ways, which influences the type of research. Not all types of research lend themselves to all questions. The major types of research used by nurse researchers appear in Table 18–1.

Research studies range from experimental to nonexperimental. Experimental research is a study in which the researcher manipulates a treatment or intervention, randomly assigns subjects to either a control or experimental group, and has control over the research situation. Because nurse researchers study human subjects, often in health care situations, a true experimental design is difficult. Consequently, nursing research is most commonly quasiexperimental or nonexperimental research. Quasiexperimental research is a type of study in which the researcher manipulates a treatment or intervention but is unable to randomize subjects into groups or lacks a control group. Nonexperimental research is a type of study in which the researcher collects data without the introduction of a treatment or intervention.

Parts of the Study

A research report begins with an abstract, which is a short summary that contains brief information about the purpose of the study, the number of subjects, the methodology used to select subjects, the type of study conducted, and the major results obtained from the study. The abstract is used by the reader to decide whether the study contains information of interest. Some nursing journals print only the abstract of a  study along with a reference that readers use to locate the full study report. An example of an abstract appears in Figure 18–3.

If the entire study is included in the journal, you will typically find the following parts:

Many of the research reports found in current nursing literature will be primary sources. A primary source is the original research report written by the researchers. A secondary source is any other published material that reports on the study. When a research report or textbook cites a study, it become a secondary source. A secondary source is an adaptation of the primary source and may contain another person's interpretation of the research.

STEPS IN THE RESEARCH PROCESS

The research process follows a series of unique and essential steps. Although these steps may vary with the type of research conducted, a typical research study will follow this order:

These steps help the researcher address a problem in an orderly way. They also help consumers of research to understand and evaluate the study. Most research articles will be organized and written according to these steps. The steps are common to all research, not only nursing research.

State the Problem

A research problem is an observation, situation, occurrence, or even a hunch that an investigator chooses to research. Initially, it may simply be an area of interest that the researcher then further defines. Most problems begin with something the researcher has observed in practice. For example, if in clinical practice a nurse observes that some diabetic patients are better able to control their blood glucose levels than others, the researcher might question the reason for this difference.

In addition to the significance of the study, researchers must consider whether it is possible to design a study to research the problem. Not all problems are amenable to research methods. For example, issues that are of a moral or ethical nature are not researchable.

Another consideration that researchers must address is the feasibility of conducting the research. All researchers face limitations, such as time, availability of subjects and resources (such as equipment), the cooperation of institutions, financial support, and the facilities needed to conduct the research. Many researchers change or limit their research problems based on the availability of resources.

The research problem or purpose of the research should be stated early in the research report. In the problem or purpose statement, the researcher should clearly identify the variables under study, the suspected relationship of these variables, and the population of interest. For example, a purpose statement might be: The purpose of this study is to compare the effects  of grief on the physiological health of women who experienced the death of a spouse within the preceding 12 months and the effects of grief on the physiological health of women who experienced the death of a child within the preceding 12 months.

The researcher's use of verbs in the purpose statement often identifies the state of the knowledge about the subject or the manner in which the researcher sought to solve the problem. The verbs explore or describe usually indicate a topic about which little is known, whereas test or compare suggests a more thoroughly developed knowledge base. Verbs such as show, prove, or demonstrate indicate possible researcher bias and should not be used.

Review the Literature

The researcher begins to study the problem by reviewing what has already been studied about it. A literature review is essential to all types of research and serves as the foundation for the research study. The researcher conducts the review by thoroughly examining all available literature related to the research problem. The purposes of the literature review are to help the researcher identify or refine the research problem, strengthen the rationale for the research, develop a conceptual framework, and provide a useful approach for the study.

Computer Searches

Computer searches are used to reach bibliographic information stored in databases. Many computer databases offer the same information found in indexes and abstracts, but the information accessed through a computer may be more current. A computer search may be completed at any computer that has access to the database. Library computers usually have access to databases, or the librarian can help to access the desired database.

Develop a Theoretical Framework

A theoretical framework is a logical but abstract structure that suggests the relationship among the variables in a research study. It allows for the organization and explanation of all of the information included in the study. Working within a framework enables the researcher to tie the research to the body of nursing knowledge. Thus, the study findings can be generalized to similar populations. All frameworks are based on key concepts and the relationships among those concepts.

For example, a researcher who wants to examine the effect of poor health on life satisfaction must recognize that both poor health and life satisfaction are abstract concepts. Such concepts are frequently organized into theories. In this example, the researcher might want to use the Roy Adaptation Model (Roy & Andrews, 1996) as the theoretical basis for the study. In this model, poor health would be viewed as a focal stimulus to trigger some kind of adaptation and lead to either life satisfaction or a lack of satisfaction. Without the model, it might be tempting to look only at the negative effects of poor health. Using a theory helps researchers clarify relationships and explain relationships. In addition, using a theory helps contribute to the testing of that theory, which increases nursing knowledge.

Identify Variables

The concepts under investigation in research studies are called variables because they are expected to change or differ from one person to another or from one time to another. The researcher is often trying to study how variation in one concept produces change  in another concept. Almost anything in people and their environments may vary and, thus, be considered a variable. If all humans were 5 feet tall and weighed 100 pounds, then height and weight would not be variables. Disease conditions are variables because not all people have the disease; those that do may have varying degrees of severity. Other examples of variables include temperature, weight, knowledge, nursing interventions received, self-concept, health, and grief.

Researchers talk about independent and dependent variables rather than about cause and effect because, even though a relationship may exist between two variables, that does not prove cause and effect. An independent variable may change during the study, but the change is expected to remain constant or to cause change in another variable. A dependent variable is expected to change with the treatment; thus, to have been caused by an independent variable. Following are three easy ways to distinguish between dependent and independent variables:

Now apply these criteria to a research problem: Do clients who receive home nursing care after discharge experience fewer complications? Home health nursing is the independent variable, because it may be the cause of fewer complications, and it is what the researcher will manipulate. Complications is the dependent variable because it is the effect. It depends on changes in the independent variable.

Clarify Operational Definitions

Research requires precision in how the concepts are being measured. An operational definition is the meaning of the concept precisely as it is being used in the study, defined in a manner that specifies how the concepts will be measured. For example, if the researcher has identified height as a variable for a study, the operational definition may be the distance from the bottom of the feet to the top of the head as measured in inches. This definition specifies that information will be collected in inches, not centimeters. Also, this definition specifies to other researchers and readers exactly how this term is used in this study, allowing for the replication of the study in the future.

Operational definitions are even more important when researchers are studying variables that are not as easily defined as height. Variables such as self-concept, pain, grief, stress, and adaptation are very difficult to operationalize. For example, in a thesis, psychosocial adaptation was defined as: "overall adjustment to health care orientation, vocational environment, domestic environment, sexual relationships, extended family relationships, social environment, and psychological distress" (Huckstein, 1995). These concepts were measured by a Psychosocial Adjustment to Illness Scale that gave specific meaning to each of the concepts. In this example, the researcher clarified what was meant by psychosocial adaptation, and how the term was to be measured. Another researcher could repeat this study using precisely the same method of measurement.

Formulate a Hypothesis

A hypothesis is a tentative prediction of the relationship between two or more variables being studied. It includes independent and dependent variables. In a previous example, the research problem was to compare the effects of grief on the physiological health of women who had experienced the death of a spouse versus a child during the preceding 12 months. The hypothesis for this problem could be: Women who have experienced the death of a child in the past 12 months will report less grief and better physiological health than women who have experienced the death of a spouse in the past 12 months. If the hypothesis fails to propose a relationship between two or more variables, it cannot be tested.

A hypothesis should be based on previous research or should be deduced from a theory. The researcher should present a sound, justifiable, logical rationale for the study hypothesis.

Select a Research Design

A research design is a researcher's strategy for testing a hypothesis. The hypothesis should guide the design, but the researcher must decide what design would be best for the study. Research designs are categorized as either quantitative or qualitative. Quantitative research is a type of study that uses variables analyzed as numbers, and qualitative research is a type of study that uses ideas that are analyzed as words. Quantitative designs usually are best suited to studies that focus on determining cause-and-effect relationships, whereas qualitative designs are best suited to studies that focus on discovery or exploration.

Collect Data

Data collection is the process by which the researcher acquires subjects and collects the information needed to answer the research question. It is the actual measurement of the study variables. Researchers use various instruments--such as questionnaires, interviews, scales, observations, and physiological measurements--to collect data.

Any instrument used in research should be reliable and valid. Reliability is the degree of consistency and accuracy with which an instrument measures a variable. For example, if an instrument is used to measure temperature, that instrument (in this case a thermometer) should measure the temperature accurately each time it is used. With physiological measures, instruments are usually straightforward. Concepts, such as health or stress, can be difficult to measure. Validity is the ability of an instrument to measure what it is designed to measure. For example, a valid measure of self-concept must measure a person's perception of the self, not another person's assessment of the study subject's belief's about self. One method of addressing validity is to ask experts to evaluate these instruments. To evaluate an instrument's validity, the reader of the research must decide if the experts are in fact experts, if the number of experts was sufficient, and if the author revised the instrument based on input by the experts.

Sampling is the process of selecting the subjects from the population being studied. It is an economical and efficient means of collecting data when use of an entire population is not feasible. Sampling techniques and the criteria used to define the population affect whether the findings can be generalized as being true of the whole population. The sample should reflect the same variations as those of the population. Generally, the largest sample size possible is the best.

There are two basic sampling techniques used in nursing research: random or probability sampling, and nonrandom or nonprobability sampling. Random sampling is the only method of obtaining a representative sample. In a randomly selected sample, each member of the population has an equal chance of appearing in the sample. It reduces the possibility of researcher bias. Nonrandom sampling involves the selection of subjects using nonrandom techniques. It is less rigorous and results in a less representative sample. The major disadvantage of nonrandom samples is that they limit the researcher's ability to generalize from the study results.

Analyze Data

The primary purpose of data analysis is to impose order on the quantity of data so that conclusions can be made and communicated. The researcher's choice of research design determines how the data should be analyzed. Quantitative research should be analyzed by numbers and statistics. Quantitative data analysis uses statistical computation to summarize the collected data, compare and contrast the data, test theoretical relationships, generalize about the population based on sample findings, and evaluate possible cause-and-effect relationships. Most researchers use computers to help with statistical analysis.

Qualitative research should be analyzed through words and logic. It depends on intuitive and analytical reasoning to guide the organization, clustering, and reduction of data. Reduction is the organization of volumes of narrative data into concepts that allow the researcher to deduce meanings. After the data are reduced, they may be displayed using tables, graphs, and matrices. From the data display, the researcher then draws conclusions and attaches meaning to the findings.

Interpret Findings

To interpret research findings, the researcher examines, organizes, and attaches meaning to the results obtained from the data analysis. Study findings should be drawn from the data analysis and related back to the theoretical framework.

The researcher forms conclusions from the current study coupled with information learned from previous research studies. When forming conclusions, the researcher must clearly state that the research supports or does not support a position; the researcher must not state that the research proves a position. Also, conclusions should result from a logical deduction of the data and not extended to include variables not addressed in the study. When formulating conclusions, the researcher should provide practical suggestions for implementing the findings in nursing. The areas of nursing where the findings can be implemented should be identified as well as implications for nursing education and further nursing research.

RESEARCH IN NURSING PRACTICE

The goal of nursing research is to improve nursing practice. Consequently, nurses must incorporate the findings from nursing research into their practices. Nursing practice identifies areas for research, and nursing research helps solve clinical practice problems. This reciprocal relationship improves practice and provides for new professional knowledge.

Evaluating Research Findings

Research in a practice profession, such as nursing, provides information needed to improve practice. For research to improve practice, researchers need to study problems that have been identified by practicing nurses. The practicing nurse needs to have the skill to evaluate research findings and develop interventions that make appropriate use of the information. Nursing research has relevance for all nurses, not just the minority of nurses who are nurse researchers. The following sections offer an introduction to the evaluation of nursing research.

Tentative Nature of Research Findings

The results of research never prove that a hypothesis or theory is true. Rather, all research results are tentative. A hypothesis tests only one small part of a theory, and researchers cannot even state that the small part of the theory tested by the hypothesis is true. The hypothesis may be faulty, the sample may be too small, the sample may not be representative of the population, or there may be another serious flaw in the study. For example, suppose that an undiscovered enzyme controls blood cholesterol levels, but researchers report that blood cholesterol levels have been proven to result from diet, exercise, metabolism, and family history. What the researcher considered proof would be false. The results of the study were not proof. They merely indicated that a relationship existed for this sample. Instead of indications of proof, watch for phrases such as the data support or the data indicate as you read studies. Research results are always tentative and are based on the sample and study involved.

Research Bias

Bias is a factor that can change or distort the results of a study. In a good research study, you can feel confident that a change in the dependent variable is because of the independent variable. However, when a researcher does not attempt to control for bias, this relationship may not be true. Bias may result from the researcher's conscious or unconscious desire to demonstrate a relationship between variables (Kelber & Pearson, 1996). Bias is commonly introduced when studies are designed to elicit specific results.

Another source of bias is the difference among subjects in groups that are being compared. When groups are formed on a nonrandom basis, the risk of bias is always present. For example, in a study in which blood cholesterol level is the dependent variable and diet and exercise are the independent variables, a major concern is that individuals with high cholesterol may differ from those with low cholesterol in ways not connected with the independent variables. Other differences, like metabolism and family history, may cause high or low cholesterol. Unless these other differences, called extraneous variables, are controlled, the resulting biases make it difficult for the researcher to conclude that lower cholesterol levels are related to diet and exercise.

When study data are collected by observation, the researcher's beliefs may unconsciously bias objective collection. To prevent bias from occurring, a double-blind technique is often used. The double-blind technique removes observer bias because both the subject and the person collecting the data are "blind" to the research objective. If this technique cannot be used, many researchers will use two or more independent observers to reduce or eliminate observer bias.

Threats to Validity

If a study has validity, it actually measures what it was designed to measure. When evaluating research, you should examine factors within the design and factors external to the study that could affect the results. These two types are called internal and external validity.

Internal validity depends on the extent to which a change in the dependent variable can be attributed to the independent variable. True experiments possess a high degree of internal validity because of the use of control groups and randomization. This enables the researcher to control for extraneous variables, thereby ruling out most alternative explanations for the study results.

Other research designs always have alternative explanations for the study results because extraneous variables in addition to the independent variable could cause a change in the dependent variable. These alternative explanations are called threats to internal validity and have been grouped into several classifications:

External validity refers to the extent to which findings can be generalized to other populations, samples, or situations. Researchers almost never conduct studies to use the findings with only one group of subjects. Researchers hope that study findings can be used by other nurses in similar situations. There are several threats to external validity, two of which will be discussed here.

Faulty Statistics

Statistical analysis allows the researcher to make quantitative data meaningful. It allows the researcher to summarize, organize, compare, evaluate, and communicate numerical data. Without statistics, data would be a mass of numbers with little or no meaning.

Statistics are classified as either descriptive or inferential. Descriptive statistics are used to describe data. Examples of descriptive statistics include averages and percentages. When researchers use statistics to make inferences or draw conclusions about a population, inferential statistics are used. The difference between the two is that descriptive statistics are concerned only with characteristics of the data obtained by the researcher, whereas inferential statistics are concerned with generalizations to a population larger than that of the data.

Scientists have developed a classification system for different types of data resulting from instruments. This classification system is important because the statistical tests performed on data depend on the measurement level. A faulty statistic can result when the researcher uses the wrong measurement level.

In addition, when researchers use inferential statistics to generalize findings to a population, the statistical tests used have assumptions that should be met in order to use the test. If these assumptions are violated, then the results of the test may be questionable. If the researcher uses an inappropriate measurement level or violates major statistical assumptions, then the study results may be invalid.

Implementing Research Findings

The responsibility for implementing research findings is shared by all nurses. In fact, nurses' behaviors and attitudes are critical to the success of any efforts to base nursing practice on research findings (Thompson, 1996). Individual nurses can contribute to implementing research findings by the following:

Expanding Professional Knowledge Base

Nursing students and staff nurses who read research reports and look for opportunities to apply sound research findings make an important contribution to the expansion of nursing knowledge in addition to improving their own clinical practice (Beyea & Nicoll, 1997). Nurse researchers are usually not interested in pursuing knowledge for its own sake but want their findings to improve practice.

When research findings are incorporated into individual clinical practice, nursing knowledge expands. Nurses identify when particular findings seem to work or not to work. From this evaluation, new hypotheses are formed, ideas are generated, and the nursing profession is revitalized. Some of the new hypotheses and ideas may never be tested in a controlled research study but, as nurses communicate new and different methods, nursing practice is improved, and the knowledge base of nursing is expanded.

Every nurse has an important role to play in making use of nursing research. Research originates from questions being asked by practicing nurses. Every nurse has not only the right but also the responsibility to ask for evidence that practices and procedures are effective. Clinical decisions must be based on sound rationales. Practices and procedures must be challenged and changed based on current knowledge rather than simply accepted because "it's always been that way."

Making Clinical Decisions

There is tremendous potential for using nursing research to make clinical decisions. The nursing process requires nurses to make many decisions. What will be assessed? What are the priority nursing diagnoses? What plan of care will produce the best outcomes? What interventions are necessary? How will the results be evaluated? When will the results be evaluated? Nursing research plays an important role at each phase of the nursing process by helping nurses make informed decisions when carrying out the nursing process.

Research-based nursing decisions begin when clinical problems are identified. If the problem has minimal significance to nurses, or if making a change or introducing a new intervention will not benefit clients or nurses, there is little point in implementing the change. However, when an important problem has been identified, you can proceed by identifying and critiquing the current research literature for information that will help you make decisions. To decide if the information is useful in solving your problem, ask questions about the transferability, feasibility, and cost/benefit ratio (Cullum & Sheldon, 1996).

Transferability simply refers to whether it makes good sense to attempt an innovation in your practice situation. Feasibility refers to the practical concerns about the availability of resources to implement the innovation. If the resources are not available, can they be obtained? The cost/benefit ratio encompasses the likely costs and benefits to various groups, including nurses, clients, and the institution as a whole. Clearly, the client is of utmost importance. The benefits to the client should be high, with minimal risk.

After determining the transferability, feasibility, and cost/benefit ratio, assess the potential for implementation for the innovation, and plan to include the innovation in your practice. Implement and evaluate the innovation. Then, make a rational decision about adopting the innovation based on your evaluation. Remember that implementing change is sometimes difficult and frequently takes longer than is originally thought.

Improving Quality of Care

As a student, the best way for you to use nursing research to improve the quality of your practice is to identify an area of nursing where you have unanswered questions. Go to your library, and conduct a literature search using both manual and computer techniques. Your reference librarian can help you locate the various indexes and provide information and assistance with computer searches. Once you have located several research articles of interest, consult the listing of periodicals for your library, and select one or two journal articles. If your library does not have research journals, the reference librarian may be able to help you obtain them through an interlibrary loan. After you have obtained the articles, it is helpful to photocopy them so that you can make notations in the margins. As you read the articles, try to identify the positive and negative aspects of the articles and any flaws in the research. Pay particular attention to the implications for clinical practice. Then identify how you can use the research to improve the quality of care you provide to your clients. Incorporating nursing research into your clinical practice provides constant renewal of your nursing knowledge and that of the profession at large.


 

BIBLIOGRAPHY

*American Nurses' Association. (1975). Human rights guidelines for nurses in clinical and other research. Kansas City, MO: Author.

*American Nurses' Association Commission on Nursing Research. (1981). Guidelines for the investigative functions of nurses. Kansas City, MO: American Nurses' Association.

*American Nurses' Association Council on Nursing Research. (1997). Directions for nursing research: Toward the twenty-first century. Washington, D.C.: American Nurses' Association.

*American Psychological Association. (1982). Ethical principles in the conduct of research with human subjects. Washington, D.C.: Author.

*American Sociological Association. (1984). Code of ethics. Washington, D.C.: Author.

*Bauknecht, K.L. (1985). Capitol commentary: NIH bill passes, includes nursing research center. American Nurse, 17(2), 2.

Beyea, S.C., & Nicoll, L.H. (1997). Research corner: Research utilization begins with learning to read research reports. AORN Journal 65(2), 402–403.

*Code of Federal Regulations. (1983). Protection of human subjects: 45 C.F.R. 46 (revised as of March 8, 1983). Washington, D.C.: Department of Health and Human Services.

Cullum, N., & Sheldon, T. (1996). Clinically challenged: Gap between nursing research and practice. Nursing Management, 3(4), 14–16.

*Goldmark, J. (1923). Nursing and nursing education in the United States. New York: Macmillan.

Goode, C.J., & Tiller, M.G. (1996). Moving research-based practice throughout the health care system. Medsurg Nursing, 5(5), 380–383.

*Hebert, R.G. (1981). Introduction. In R.G. Hebert (Ed.), Florence Nightingale: Saint, reformer, or rebel? Malabar, FL: Robert Krieger Publishing Co.

Huckstein, M.E. (1995). The relationship between hardiness and adaptation to lifestyle changes in multiple trauma patients. Master's thesis. Miami, FL: Florida International University.

*Kalisch, P., & Kalisch, B. (1978). The advance of American nursing. Boston: Little, Brown & Co.

Kelber, S.T., & Pearson, B.D. (1996). Translating research into practice: Ways bias may occur in research. Journal of Urological Nursing, 15(1), 1216–1221.

*National Center for Nursing Research. (1993). Research priorities from the National Center for Nursing Research. Bethesda, MD: Author.

*National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1978). Belmont report: Ethical principles and guidelines for research involving human subjects. Washington, D.C.: U.S. Government Printing Office.

*Nightingale, F. (1859). Notes on Hospitals. London: Longman.

Polit, D.F., & Hungler, B.P. (1995). Nursing research: Principles and methods (5th ed.). Philadelphia: J.B. Lippincott Co..

*Rogers, B. (1989). Florence Nightingale and research. AAOHN Journal, 37(6), 238–239.

*Rothman, D.J. (1982). Were Tuskegee and Willowbrook studies in nature? Hastings Center, 12(2), 5–7.

Roy, C., & Andrews, H. (1996). The Roy adaptation model (2nd ed.). Norwalk, CT: Appleton & Lange.

Thompson, D.R. (1996). Getting research into practice. Intensive and Critical Care Nursing, 12(4), 191–192.

*Vessey, J., & Gennaro, S. (1994). The ghost of Tuskegee: The Tuskegee study of untreated syphilis in the Negro male. Nursing Research, 43(2), 67.