Small sample studies: unique contributions for large sample outcome studies.

There are many important issues relevant to the course of illness and outcome of chronic psychiatric disorders that are not readily amenable to investigation with existing large sample, quantitative methodology. For example, complex interactive phenomena that change over time, such as the longitudinal impact of changes in social and work function on levels of symptomatology, are particularly difficult to quantify and assess in large n studies. Small sample studies are advantageous because they offer the ability to examine the longitudinal interplay of a wide range of variables in individual patients. The small n approach is useful for delineating important new course variables, developing models of course change, and generating unique research hypotheses. In this article, small sample "qualitative" methodology is reviewed, and data derived from recent small n, intensive longitudinal studies are presented. Methodological issues related to addressing qualitative data in large sample, controlled studies are addressed, and specific research hypotheses generated by small sample studies for consideration in future large sample outcome studies are suggested. It is concluded that small sample studies offer unique contributions to the understanding of the determinants of course of illness.