Abstract
This article provides an introduction to a number of statistical techniques for investigating the structure of measures of personality. Each of these techniques is illustrated through its application to a set of 12 items from Snyder's (1974) self-monitoring scale. Approaches are presented for both continuous and dichotomous observed measures and for underlying constructs that are assumed to be continuous as well as those that are assumed to be discrete classes. Our emphasis is on techniques which permit tests of hypothesized structures (with a special focus on confirmatory factor analysis); however, exploratory approaches are also considered. We identify the critical assumptions, limitations, and areas of application of these techniques with an aim toward fostering their critical and appropriate usage.
View more >>