A re-analysis of the Situational Attitudes Scale (SAS) was conducted to investigate lts generality, validity and reliability and to suggest a more conservative method of analyzing the data. Three forms of the SAS were developed under the name Situational Attitude Inventory. Forms A and B corresponded to the originals; they tested attitudes toward whites and blacks, respectively, in varying social situations. Form c was developed to test the validity of the social situations for differentiation among alternative referents, which included welfare, age, ethnicity, and physical characteristics. The SAS was administered to a random sample of new freshman and transfer students at Michigan State University in September, 1970. Data were analyzed using factor analysis and multiple t-·tests of individual items and multivariate analysis of variance (MANOVA) of scales constructed by the method of reciprocal averages (RAVE). Factor analysis demonstrated independence of SAS situations. Multiple t-test data showed that white 1970 new students had more negative attitudes toward blacks than whites. Profile analysis revealed that students reacted to the scales differentially, depending on the form of the instrument completed; aspects of the social situations affected their responses as well as the particular descriptive characteristics assigned to the actor(s). There is a "saving" of at least 89 statistical tests using this method of analysis over the previous one, as well as having a more widely understood estimate of reliability.
Available at: http://works.bepress.com/charles_g_eberly/19/