This proposal involves the development of research methodology and the accompanying technology. A research design that is being used more frequently in practice-based healthcare research is the single subject research design (i.e., clinical trial of N=I). However, the statistical assessment of clinical effectiveness of the intervention when using this design is controversial. The data analytic approaches most often chosen for this assessment have not been shown to be valid. The use of a nonparamedic smoother has been validated as a descriptive tool, but further development and testing needs to be completed for an inferential tool. This proposal involves the development and validation of a general data analytic strategy for the statistical assessment of data from a single-subject research design. Of focus is the A-B research design. It is proposed that such a data analytic strategy might combine a nonparamedic smoother with a randomization test, thus resulting in a descriptive and inferential means of examination and the determination of effectiveness of the intervention for the assessment of practice-based single-subject designs. The development of the methodology, the empirical investigation of this methodology, the application to ongoing medical research, and the needed technology for this design is proposed. It is anticipated that, once developed, this methodology and technology will be available to the health sciences research community.
| Janosky, J E (2005) Use of the single subject design for practice based primary care research. Postgrad Med J 81:549-51 |
