Methodological considerations involved in studying the impact of the sheltered workshop setting on the behavior of psychotic patients are discussed within the framework of a pilot study designed to examine the effects of work-flow variations on interaction in small work groups. An observational technique for measuring behavior through systematic sampling and the problems associated with analysis of quantitative behavioral data are presented. Analysis of the pilot study revealed that increases in the amounts of interaction among schizophrenics working in a ward-sheltered workshop may result from relatively simple variations in the ways in which the jobs are structured.