These studies, conducted at an experimental secondary school which sought to create "community felling" by not separating students by grade or ability, looked at the students' social patterns and groupings and at the impressions which subgroups had of own group and other groups. The first is based on an all-school sociometric questionnaire; the 2nd, on interviews with members of friendship groups that were exposed in the sociometric. It was found that there were some readily identifiable friendship groups, and that they were more homogeneous than diverse, there were many isolates who received no choices at all; there were few cross-sex or cross-grade-level friendship choices; and choices within the Home Groups showed great differences from the choices made in the school at large. There were few cross-race choices. Students in some of the clear-cut friendship groups frequently lacked information about others, had negative feelings towards them, and spoke of groups not identifiable in the sociometric in terms of stereotypic language. Particular efforts were made to minimize differences amongst the individuals within their own groups. (KS)