Construct Validation of a General Social Attitudes Scale (SAS_G)

Abstract

In this article, we describe the construct validation of a General Social Attitudes Scale (SAS_G), designed to measure the basic dimensions of social attitudes in both self-reports and peer-ratings. A large sample of Croatian university students (N = 452) used the SAS_G to describe their own social attitudes, which were also described by 452 of their acquaintances using the same instrument. All SAS_G subscales showed reasonably high internal consistency reliability estimates as well as appropriate convergent and discriminant validity based on self/peer correlations. Separate Confirmatory Factor Analyses (CFA) of both self-reports and peer ratings yielded five replicable factors and acceptable indices of fit. However, the validation analyses against lexically based Saucier’s (2008 ) ISMS instrument showed little convergence, indicating that these two approaches to general social attitudes scale construction can lead to the identification of different basic constructs.