Research issues for clinical designs

Author(s): Berman, B.

Abstract: The purpose of this paper is to raise the consciousness of researchers in the complementary medical community to the question of appropriateness of research design and some of the questions that must be asked before choices are made. The goal of any experimental design should be to increase primary variance, decrease secondary variance, and avoid confounding. Research questions should be asked based on thorough reviews of the current literature and the research design built around this information and the practicality issues that are salient for clinical research. For example, randomization should not always be chosen as the mechanism for achieving group equivalence in multigroup designs. lf the sample size is very small - < 20 per group or if patients are unlikely to agree to randomization, for example, other techniques may be used effectively with a priori planning and post hoc statistical control. Additionally, the paper indicates some of the advantages of the use of a quasi-experimental (outcomes) design when randomized clinical trials are not practicable.

Zurück | Weiter