What is expected in terms of power calculations and sample size estimates for the formative studies, early-phase trials and pilot studies to be included in this RFA?
A. The issue of powering early phase behavioral intervention trials is complex and there is much debate about the best method(s) to use. Clearly sample size estimation for the studies involved in this RFA will require some creative approaches as standard power calculations are unlikely to be appropriate. Despite the challenges, we think it is important to consider how many subjects are needed to be able to derive meaning from the data you collect at each step. As such, power needs to be considered in concert with an estimation of criteria or milestones met. For example: To consider the intervention successful enough to proceed to the next phase of research, how many participants are needed to detect “x” magnitude of response? For formative research, you usually proceed to “saturation” point, where no new information is provided by subsequent participants. You might consider consulting a statistician familiar with early phase biomedical trials as they will likely be familiar with the ex
