Sometimes Less Is More: On Sample Size in Medical Research

Large samples have well-known merits in empirical research where multiple factors are at work, some beyond control. Large samples also provide a precise estimate of the effect size, although they produce results for groups and not for individuals. Nonetheless, there are situations where intensive, in-depth, study of a few may be rewarding, particularly in medical research that focuses on basic mechanism rather than treatment effect. With extreme care, logical structure, advanced tools, and rigorous reasoning, many sources of variability can be minimized or eliminated. Thus, small samples tend to foster brilliance by focusing on improved decision making. Reduced cost, faster completion, and clinically relevant effect sizes are among other advantages. Many successful studies have been done on small samples. These merits of small sample studies in certain types of medical research deserve serious consideration.