Abstract
Group construct systems were defined through application of a group construct grid that elicited types of groups in which an individual had participated or closely observed. The Group Grid was administered at the beginning and the end of a 4-month graduate class. The meanings of groups to participants were interpreted by examining changes in Group Grids over time and in relation to high and low levels of group performance. The meaning of individuals to the group was assessed by peer evaluations given at the end of the project. High and low levels of peer evaluation were then predicted by an individual's Group Grid measures. The most consistent findings were that greater variation in application regarding types of groups and characteristics about groups showed strong associations with both high levels of group performance and high levels of peer evaluation. Suggestions are given for future research