For their study, Riegel, Dickson, Goldberg and Dietrick (2007) used a mixed-methods design, that is, a study design that combines both qualitative and quantitative methods. For the qualitative part, interviews were conducted with the patients with the use of open-ended questions and probes. Transcripts of the interviews were analyzed, and one of the investigators classified each of the participants as “poor,” “good” or “expert” in heart failure self-care.
For the quantitative part, the participants were asked to complete a sociodemographic survey, two neurobehavioral test, and questionnaires measuring the following variables: symptoms and physical limitations; heart failure self-care; excessive daytime sleepiness; family functioning; and depression. The quantitative data obtained was compared between groups of different heart failure self-care expertise using ANOVA and ch-square analysis through SPSS 13. 0. The Appropriateness of the Design in Relation to the Research Question
Riegel and colleagues (2007) state, “In this study, the development of HF self-care expertise is described” (p. 235). They further wrote, “To understand HF self-care, we explored ways that available information shaped decision making about self-care, presumably through its influence on knowledge, experience, skill, and values” (pp. 235-236). To the extent that Riegel et al. (2007) wanted to “describe” and “understand” heart failure self-care, the mixed-methods design was well-chosen.
The combination of quantitative and qualitative data gave them the ability to use the results from one method to clarify, expand or explain the results obtained using the other method. For example, the participants’ scores on the heart failure self-care questionnaire were better understood in terms of the specific ways they managed their condition as they related to their interviewer. On the other hand, the variables measured quantitatively by the questionnaires and tests provided an objective comparison between the participants classified as “poor,” “good” or “expert” after an analysis of their interviews.
Threats to the Internal Validity of the Design Of the six internal validity threats, three were controlled for, two were not controlled for, and one was not an issue in this study by Riegel and colleagues (2007). History. This was not controlled for by Riegel and associates (2007). While it is true that personal or social events outside of the variables being studied could affect the patients’ expertise in heart failure self-care, this is less important in a study that does not seek to establish cause and effect.
Moreover, this threat is greater when data collection is performed two or more times, however, in this study, data collection was done only once for each participant. Selection. This was definitely not controlled for in this study. Riegel et al. (2007) admitted that “extreme case sampling was used to identify 29 chronic HF patients predominantly poor or particularly good in self-care. ” This meant that their subsequent labeling of each participant as either “poor,” “good” or “expert” in heart failure self-care was already affected by their decision to limit their participants to those on extremes.
Their sample was not reflective of the entire continuum of heart failure self-care expertise that undoubtedly exists in the entire population of patients with heart failure. Maturation. This is not an issue in this study, as data collection was done only at one time for each participant. Mortality. Riegel and colleagues (2007) do not report a single participant dropping out from their study. The researchers even made an effort to control this by offering their participants cash incentives for completing the study. Testing and Instrumentation. This was controlled for by Riegel et al. (2007).
The variable of prime concern in this study was heart failure self-care expertise, as measured not only by a single questionnaire but by the participants’ self-reports on the steps they take to manage their condition. Furthermore, Riegel and colleagues (2007) report, “Standardized instruments were used to quantify the level and type of self-care in which the patients engaged and to gather data on factors anticipated to influence self-care. ” Temporal ambiguity. This is usually an issue in correlational studies, where the cause and effect relationship of the independent and dependent variables are unclear.
However, on analysis, it would be difficult to conceive how the dependent variable – heart failure self-care expertise – can cause the independent variables, such as excessive daytime sleepiness. Thus, this threat is relatively controlled for in this study, not by design but by the researchers’ choice of variables. Threats to Construct Validity Of the five threats to construct validity, two were controlled, one was not controlled, and two were not considered as issues in this study. Reactivity to the study situation.
A desire to be seen in a good light might have influenced the answers that the participants gave, both in the interviews and in the questionnaires. This is a problem that plagues all studies relying on self-reports and was not controlled for by Riegel and associates (2007). Researcher expectancies. This was controlled by Riegel and colleagues (2007) by assigning just one investigator – who did not know the results of the quantitative investigations – to classify patients’ expertise in heart failure self-care. Novelty effects.
Since the variables studied pertained to activities or situations that were already operational before the study was conducted, they could not have been affected by this threat. This is therefore not an issue in this study. Compensatory effects. Riegel et al. (2007) controlled for this threat to construct validity. Since the participants were only classified after the data collection, there was no way compensatory rivalry or compensatory equalization of treatment could have taken place. Treatment diffusion or contamination. This is not an issue in this study because there were no different treatments between groups.
All participants underwent the same process, and groups according to expertise were only identified after data collection. The Study Redesigned Inasmuch as Riegel and colleagues sought primarily to “describe” heart failure self-care expertise, the mixed methods design appears to be the most suitable for this study. The qualitative aspect provided the depth necessary for adequate understanding of heart failure self-care. However, this provided an avenue of subjectivity – namely, that of the investigator in charge of classifying patients according to expertise.
I would probably use in-depth interviews with a pilot group of patients as a preliminary data-gathering tool. Out of these interviews, I would compile a list of knowledge and behaviors that comprise heart failure self-care, and I could then make a questionnaire to be used in the study proper. This would ensure objectivity of expertise classification. Just as importantly, I would increase the sample size and enroll participants that are representative of the entire population of heart failure patients.
This would ensure that I do not get lopsided results, and it would increase the study’s statistical power. Only then can the results truly be generalizable to all heart failure patients. Nevertheless, the study provided valuable insight to the process of acquiring expertise in HF self-care and can be the platform from which to launch more comprehensive studies. References Riegel, B. , Dickson, V. V. , Goldberg, L. R. , & Deatrick, J. A. (2007). Factors associated with the development of expertise in heart failure self-care. Nursing Research, 56(4), 235– 243.