Exact Test Size and Power for Small Samples Using an Internal Pilot Study

Marlar and Welsh (2001) proposed a prospective ascertainment strategy for a case-control study of deep venous thrombosis with a binary exposure. This study design relied on prevalence estimates of certain genetic polymorphisms from previous studies using different populations. They planned on collecting a portion of the data to use as an internal pilot in order to better estimate these prevalences using their population of interest. The internal pilot data were used to estimate the population prevalences and recalculate the required sample size. All the data, including the internal pilot data, were used in the final analysis. They used a x(2) to analyze the resulting 2 x 2 table. We have been unable to find a paper in which the authors considered power and sample size for a case-control study with an internal pilot study.Air Force Inst of Tech Wright-Patterson AFB OH is the author of 'Exact Test Size and Power for Small Samples Using an Internal Pilot Study' with ISBN 9781423506287 and ISBN 1423506286.