For this issue's column, I would like to begin a discussion of quantitative nursing research. Quantitative designs are those that translate information from observations and assessments into numerical data. The data are then analyzed statistically according to a preestablished plan. Let me reemphasize preestablished and plan. The types of quantitative research include descriptive, correlational, quasi-experimental, and experimental. Descriptive research explores and describes phenomenon and helps to identify relationships. Correlational research, as the name implies, looks for relationships and the strength of relationships between variables. Quasi-experimental research examines cause and effect in nonrandomly selected samples when variables cannot be completely controlled by the researcher. Experimental research, highly controlled and systematic, studies cause and effect, comparing randomly assigned control and experimental groups (Burns & Grove, 2009, p. 700).
Despite slight variations in formatting from journal to journal, all quantitative research designs follow a similar formula. So, when embarking on any quantitative study, the researcher should first perform an in-depth literature review based on the question to be answered and then write a research proposal. A well-crafted proposal serves as a blueprint for the study. Elements of the proposal include a statement of the research problem. The problem is justification for why this study is necessary. Also, it should include the significance for the profession and the background. Contained in the background are a few key studies that have been performed related to the research question (Polit & Beck, 2017, pp. 74-76).
After the problem, a narrow, focused aim or purpose statement should be written. As previously described (Roe-Prior, 2022), elements of the purpose statement are the population to be studied, the variables of interest, and the setting. The purpose statement should alert the reader to the study design by beginning with a phrase such as to compare groups, to describe a phenomenon, to establish a relationship, or to determine the effect of a treatment or intervention (Polit & Beck, 2017, p. 75). Although only one sentence, the purpose statement provides a great deal of information and can serve as a guidepost for the researcher.
Next, a brief review of the relevant literature is provided. Through this focused, comprehensive, and up-to-date review of the relevant empirical and theoretical studies, the researcher demonstrates her knowledge of the topic. Generally, any studies cited should have been performed within the last 5 years, although essential landmark studies may be cited. Although all research is guided by some theoretical framework, not all are explicitly expressed. A framework is an abstract theoretical link to nursing's body of knowledge. When identified, a theoretical framework should be clearly described and linked to the study purpose. Concepts and how they are defined for the research purpose must also be provided.
When performing any study, it is important to conceptually and operationally define your variables. Some, like physiological variables, require no additional conceptual definition but may require operational definitions if being measured in a unique way. Although all quantitative research has dependent variables, not all has independent variables. Research variables are concepts that are measured, manipulated, or controlled. A dependent variable is the variable the researcher wishes to predict or explain, or the variable that changes after a change in another variable, or the effect in a cause-and-effect relationship. It is sometimes referred to as the outcome variable. The independent variable precedes the change in a dependent variable or causes the change in the dependent variable. It is the one manipulated by the researcher and sometimes called the treatment or experimental variable (Burns & Grove, 2009, pp. 696, 703, 720).
An example I would like to use is one performed by myself and my colleagues. One of the aims of our pilot study was to determine the factors contributing to baccalaureate nurses (BSN) leaving or planning to leave hospital nursing (DiMattio et al., 2010). A pilot study is one in which the feasibility of a larger study is determined. From the review of the literature, several variables were identified that predicted intent to leave for nurses of any educational level. Up to that point, no studies could be found which looked at only BSNs' intent to stay at the bedside, establishing the significance of performing the study. Because we were specifically interested in bedside nursing, our conceptual definition excluded nursing in advanced practice or administration. Job satisfaction, an important factor in nursing retention, had been shown to be influenced by the practice environment. The practice environment was defined as the "organizational characteristics of a work setting that facilitate or constrain professional nursing practice" (Lake, 2002, p. 178). Because we used Lake's definition, with her permission, one way we measured job satisfaction and, hence, intent to stay was with her Practice Environment Scale, a 31-item, five-subscale instrument derived from the Nursing Work Index. The instrument evaluated nurse participation in hospital affairs, nursing foundations of quality care, nurse manager ability, nurse leadership and support for nurses, staffing and resource adequacy, and collegial nurse-physician relationships (Lake, 2002), thus operationalizing intent to stay.
In the study cited (DiMattio et al., 2010), which was a cross-sectional, descriptive-correlational study, we used a survey to collect our data. Hence, no variable was manipulated; therefore, there was no independent variable. If, on the other hand, we had instituted a nurse residency program, which we then provided to half the new hires and compared the two groups on intent to stay, our independent variable would be the residency program, and as before, the dependent variable would be intent to stay. The change in design would also dictate a change in the statistical test used to analyze our data. But I'll save that discussion for another day.
Another important consideration is extraneous variables. These may be controlled or uncontrolled, recognized or not. If it is a variable that cannot be controlled or is only recognized once the study is underway, it is known as a confounding variable (Burns & Grove, 2009, p. 178). If a variable cannot be controlled or is unrecognized, it may interfere with the validity of a study's findings. Thus, a researcher should attempt to identify and, if possible, control all the variables that may influence the results of a study. For example, if our residency program involved various preceptors, a standard preceptor orientation provided prior to the institution of the residency program would minimize any difference attributable to individual preceptors and control that extraneous variable, as much as humanly possible. During the course of the theoretical study, had it been recognized that the control group preceptors were more involved in teaching the orientees than those in the intervention, or residency group, then any inability to find a difference on intent to stay between the two groups would be confounded by the control group preceptors' behavior. Any variable that can be controlled should ensure, when comparing two groups, that they are as similar as possible demographically and diagnostically at the start, so that any differences at the study's end would not be attributable to initial group differences. The goal is not to be surprised by confounding variables but to anticipate and plan the study in such a way to minimize their effects. However, when dealing with human beings, there is always something.
Although the preceding may be old hat to many of you, where I have seen some researchers struggle is in defining their variables conceptually and then translating them to a more concrete or operational definition. Remember, a conceptual definition provides the theoretical meaning of a concept and is based on the framework of the study. An example provided was intent to stay at the bedside. A variable may be operationalized, or transformed into a measurable variable by defining it concretely or by using a valid and reliable instrument with which to analyze an abstract concept-in the above example, the Practice Environment Scale. Be theoretically consistent, stick to your purpose, and know what you are trying to measure. Identify and minimize the effects of extraneous variables and find a valid and reliable instrument with which to measure your variable(s), do so in a consistent and methodical manner (more on Methods in the next issue), and make certain to obtain institutional review board approval and participant agreement prior to any human subjects research. Oh, and Good Luck!
References