Very often nurses in practice settings ask questions that have both real-world (applied) and theoretical (basic) research components. Although the initial research question is usually restated several times before a researchable question is developed, it is possible, as this article describes, to retain the realworld quality of your research while contributing to theory as well.
The project in this article was of a long-term care gerontological nature; however, the basic process of obtaining and utilizing data has widespread application across many care settings. For this project we began with a practical concern about using simple counts of the number of ADL dependencies as a reliable indicator of level of dependency in daily living. This question evolved into a basic research problem regarding the theoretical basis for counting the number of ADL dependencies that had been common practice in research and practice for the last 25 years.
We were very fortunate to be in a state which has a computerized information retrieval and data storage system (Virginia Medicaid Long Term Care Information System). Coincidentally, this screening data included the exact ADL dependency criteria needed for the study question. This was a critical issue in our decision to use an existing data base since data in some other form may not always be easily translated into the theoretical framework of interest.
Data Access - The first step in securing access to the data is preliminary contact to determine if the agency has any interest in allowing access to the data base. This can generally be facilitated by previous contact with the agency and/or some acquaintance with individuals responsible for the data system. Preliminary contact may simply be a telephone call expressing interest in accessing the data and requesting permission to proceed with a study proposal to the agency staff. In our case, we also followed the written proposal with a face-to-face meeting with involved agency staff to answer questions or concerns generated by the written proposal.
Meeting With Agency Staff- A face-to-face meeting, regardless of a strong-felt need to do so, is suggested for two reasons. First, the meeting humanizes and personalizes the research process. We were located in a distant comer of the state opposite from the state agency. Our main method of communication was by long distance telephone. By arranging a face-to-face meeting, everyone had the opportunity to put a face to a voice on the telephone. When problems arose (and they did) we found it much better to speak to someone we had met in person.
A meeting also allows each party an opportunity to ask questions and clarify even minor issues in the context of relatively informal discussion. It allows the researcher to expand on certain concepts that perhaps he or she did not want to belabor in a proposal and it allows the agency staff to discuss areas of interest and research questions they might have in conjunction with the proposed research.
It is important to listen to the agency staff and attempt to include, if possible, their concerns in the research design. This is often easily accomplished with minor changes in the original research design. If minor programming changes will generate answers for the agency, this seems a small gesture of appreciation for accessing the data.
Tape Specifications - Once you have permission to access the data you must be able to give the agency the appropriate tape specifications to produce the data tape as well as correctly identify the variables of interest and sampling method. This information can be obtained from your computer center or department where your data tape will be mounted for data analysis.
Variable Transformations - Our dependent variable (ADL dependency) was easily translated from the data set to our study variable. The independent variables, however, were not always in the form we would have used under perfect study conditions. For example, we had documentation of up to six medical diagnoses, but we were not able to determine severity of the condition from the information given.
Advantages/Disadvantages - One major advantage was, of course, the comprehensive nature of this statewide data set. We used 3611 cases in the study. Collecting new data on this number of cases statewide would have been a time-consuming, costly undertaking. In addition, the quality of the data would probably not have been any better and perhaps less reliable than the screening data which were collected by healthcare professionals.
The major disadvantages were the limits on generating independent variables of interest because of the intrinsic limitations of the data set. While we did develop an explanatory model for our research, we very often said "if only we had information on (so and so)." The operationalization of the group of variables we termed "social environmental variables" probably suffered the greatest simply due to lack of detailed information in the screening data.
Results/Reports - When data analysis is completed and it's time to record the results, two reports are probably in order. The first report, the traditional research report, generally includes a complete description of the research method, statistical procedures, results, and discussion. For a basic research problem, theoretical considerations would be emphasized. The second report to the agency should be practical and interpretive. This report very often enables the researcher to express the results in "real-world" terms. It can be an exciting process for making practical sense of your research.
Agencies and organizations have a wealth of information for basic and applied research. Unfortunately, this data source is often overlooked for nursing research questions. As researchers and practitioners begin to work more closely to solve the problems and concerns in health care, the research opportunities in practice settings will continue to expand and grow. The process for obtaining and utilizing existing data sets, such as our project, could facilitate mutually beneficial research projects for nurses as well as the interdisciplinary team.
- George L (Ed.): Symposium: Survey research in the study of aging: Possibilities and problems in the use of archival data. The Gerontologist 1979; 19(2):196-219.
- McArt E, McDougal L: Secondary data analysis - A new approach to nursing research. Image 1985; 17(2):54-57.
- Staff: Data resources for research on aging. Bulletin of the National Archives of Computerized Data on Aging, Spring 1984.