"An exploratory study of four residences for mentally retarded people and six other housing settings for nondisabled people explored two hypotheses: (1) that differences exist between institutional and other residential environments which are measurable; (2) that architectural differences which exist between institutional and other residential settings can be related to differences in expected behavior. The first hypothesis was supported by a preliminary bedroom measure for architectural characteristics which indicated a polar relationship between the descriptors "institutional" and "homelike." Drawings and slides of environments tested among students were found to be consistently categorized, indicating shared cultural expectations for behavior."