evaluating published research: discussion/conclusions
TRANSCRIPT
( RESEARCH FOR THE CLINICIAN J Evaluating Published Research: Discussion/Conclusions
P ublished research articles usually follow a standard
format: a literature search; identification of the problem; a statement of purpose; descriptions of the research sample, methods, and equipment; and a discussion or conclusion section. * Each section builds on the previous one, providing an evolutional continuum that follows through to a logical finale. The final section of a research article provides important information for readers and should be carefully considered when evaluating the study.
Augmenting the more technical results part of the article, the discussion or conclusiont section should accurately and without bias frame the statistical results in an appropriate and nontechnical context. In addition to interpreting results, this section should clearly relate the re-
*When an abstract is included, a section of text dedicated to summarizing the research study may not be required, depending on the journal's editorial philosophy.
t"Discussion" and "conclusion" may be used interchangeably when describing the final text section of a research paper. It is unusual to find both discussion and conclusion sections in a published article. Preference for one term over the other is often at the discretion of the editorial staff of a publication.
sults of the study to the original research question, thereby completing the investigative loop. Authors may also explain why they think their study turned out as it did, and they may point out how it differs from or agrees with previously published research. This is especially true when the results obtained differ from expectations. Limitations of the study are often included in a discussion section, as are suggestions for future study.
Unfortunately, authors sometimes take too much latitude in interpreting their results, leaving readers to wonder if they completely misunderstood the technical results or misread the purpose of the study. An often-told witticism beautifully illustrates the problem of erroneous conclusions:
A researcher teaches a frog to jump on command. He then anesthetizes one of the frog's legs and tells the frog to jump. The frog jumps. Next, the researcher anesthetizes a second leg and again commands the frog to jump. Again the frog jumps. This scenario continues until all four legs of the frog are anesthetized, when, upon command, the frog does not jump. Whereupon, as a result of his study, the researcher concludes that the frog is deaf.
When evaluating published research, it is important to compare carefully the technical results of the study
246 JOURNAL OF HAND THERAPY
with the authors' stated conclusions. If the two do not match technically or logically, then readers should question why. Whether unintentional or intentional, subtle or blatant, author bias is inappropriate in scientific reporting. Incongruity between results and conclusions is usually identified in the review and vetting processes of professional journals. However, bias can and sometimes does defy the scrutiny of reviewers and editorial staff, especially when the particular bias is a commonly accepted tenet. Ultimately, it is the consumer, therapist or surgeon, who must decide the appropriateness of the authors' conclusions. This is not a task to be undertaken lightly, for research conclusions often directly influence patient treatment.
Understanding the purpose of each section of published articles is critical to improving hand therapy research and clinical investigation. The identification of consistencies and inconsistencies between the results and the discussion/conclusion sections of a research article is but one of the many integrated methods for assessing the validity of published studies.
Elaine Ewing Fess, MS,OTR,FAOTA,CHT
Correspondence and reprint requests to Elaine Ewing Fess, MS, OTR, FAOTA, CHT, Hand Research, 65 Eagle Creek Court, Zionsville, IN 46077.