Research Design & Stats
Methodology of Data Collection
Protocols and procedures establish the reliability and reproducibility of the study’s results. They set forth who will collect the data and how they will be trained. Have a written protocol for how data is collected, then follow the same procedure all the time. Do not be tempted to make “adjustments” to the protocol as the study proceeds even if better ways of operating are identified. Without adherence to the study’s protocols and procedures, the data will be meaningless. Any change in the protocol invalidates the results because inconsistency in how they are attained destroys their dependability (Piantadosi, 1997, Chap. 4).
The person collecting or analyzing the data should be “masked” or “blinded” to the identity of the groups being compared. Masking is an important aspect of data collection and analysis in experimental research so that the outcomes are not biased by the researcher’s expectations (Friedman, et al.,1998, Chap.6). All the data must be collected in a dependably consistent manner by a blinded investigator to obtain a reliable answer to the study’s research question.
Statistical Analysis and Inference
Data analysis is always done statistically. The research question and experimental design determine in advance of data collection how the data will be analyzed statistically. However, the process cannot be entirely anticipated until the data is available for analysis. In the conventional comparative research design using parallel groups, initial statistical tests comparing the two primary groups of interest are made even if the groups are stratified by other factors such as gender or age. If an overall, or main effect, is detected between the main groups, then the subgroups can be tested post hoc to determine the exact location of the effect within the study population’s strata.
The limitation of this type of study design is that interaction between effects cannot be determined. This is a problem particularly if the effect in one subgroup runs counter to the overall, main effect of interest such that the effect in the subgroup negates the main effect. The factorial study design adjusts for this possibility by allowing statistical analysis for main effects as well as the interaction of effects (DeMuth, 1999).
From statistical analysis, inferences can be made about the results of the study but to have meaning, the results of the study should be interpreted in light of the current times and the state of the art and what it may mean for the future. If an effect is detected by the study, then the various possible explanations of the effect can be considered. Interpretation of why or how an effect may have occurred is an issue that can be covered within the “Discussion” section of the published report.
Reading and Interpreting The Life Care Planning Literature
Ultimately, the professional should be prepared to critically evaluate a study reported in the literature and apply that new knowledge to their professional practice and future investigative endeavors. The reader should read and interpret the published literature in the professional field to determine for his/herself what to take away from the reading rather than accepting carte blanche the conclusions presented by the research investigators.
The research concepts presented in this text offer a foundation for developing a general appraisal. Some critical questions a reader might ask of a published report include:
*How does the research design, methodology and statistical analysis affect outcome?
*Has the study been powered appropriately?
*What is the period of observation? If comparisons between LCPs were made after 5 months and the updated LCPs were completely congruous with the original LCPs, the meaningfulness might be suspect. However, if the window of observation were extended to 5 years, and the same result obtained, it might be meaningful.
*Was the research protocol appropriate and was adherence to it stringent?
*How credible, ethical and moral is the investigator?
*Does this researcher hold reputable credentials and is the work supported by a research institution?
*Has the author addressed the points that a study needs to address?
Elwood (1998, Chaps. 10-15) presents guidelines for critically evaluating and interpreting specific types of research studies.
The purpose of the published literature is to inform the public in general and future research efforts in particular. Therefore, it is paramount that future investigators and readers are equipped to read the work and parse out for themselves the strengths and weaknesses of the research, to enable the specialty practice of Life Care Planning to move forward into the future.
Developing a Research Study Proposal
Inexperience does not preclude a professional from entering the arena of scientific research. All investigators were once novices. In developing a research study proposal, seasoned mentors serve an important function for experienced as well as novice investigators. They can guide the researcher through the investigative process and give key direction. Sometimes the key direction is an introduction to the person who can help at a particular impasse. Sage mentoring is an indispensable resource to support the professional who embarks upon the scientific course of investigation.
A second essential resource to facilitate research efforts is database accessibility. Access to larger caseloads may be obtained through research institutions such as The Foundation for Life Care Planning Research. This foundation’s purpose is to help develop research proposals, provide opportunities for mentoring, and access to significant databases. For the professional who is developing a research study proposal, The Foundation for Life Care Planning Research offers important support.
Suggestions for Further Reading: Instrumentation and Data Collection
Bourque, L., & Fielder, E. (1995). How to conduct self-administered and mail surveys. Thousand Oaks, CA: Sage Publications.
Buros Institute. (2003). Mental measurements yearbook (15th Ed.). Buros Institute of Mental Measurements.
Fink, A. (1995). How to analyze survey data. Thousand Oaks, CA: Sage Publications.
Fink, A. (1995). How to design surveys. Thousand Oaks, CA: Sage Publications.
Fowler, F. (1993). Survey research methods. Thousand Oaks, CA: Sage Publications.
Frey, J., & Oishi, S. (1995). How to conduct interviews by telephone and in person. Thousand Oaks, CA: Sage Publications.
McLaughlin, P. (1990). How to interview: The art of making questions. North Vancouver, B.C.: International Self Counsel Press.
Stourthamer-Loeber, M., & Bok Van Kammen, W. (1995). Data collection and management: A practical guide. Thousand Oaks, CA: Sage Publications.