Three interrelated processes; storage, retrieval, and judgment, are distinguished in an attempt to describe stereotype formation. The results of four experiments are described. These experiments were designed to answer some of the questions about what information is learned, how it is stored, how it is retrieved, and, once retrieved, how it is used to form stereotypes about group characteristics. Experiment 1 attempted to show that subjects organize the information they learn about individual group members in different ways, resulting in varying judgments about the characteristics of the group. This result depends in large part on mnemonic organization as it relates to demands on memory. Experiment 2 was designed to see whether subjects would estimate the frequency of individuals with extreme physical attributes to be greater than the frequency of equal numbers of individuals with less extreme physical attributes. Experiment 3 sought to discover whether the same overestimation of extreme instances would occur when using social rather than physical stimuli. Experiment 4 was designed to examine subjects' recall for instances consistent and inconsistent with their original expectancies. (Author/MC)