Evaluating a measure of the five-factor model of personality

An evaluation is made of Goldberg's (1992) 100 Unipolar Markers of the five-factor model of personality. The factor structure of these items in samples of older men from the Normative Aging Study and undergraduate students are examined, and both item transformation and consistency testing approaches are used to evaluate replications of the five-factor structure. Results show that the five-factor structure is difficult to replicate in the sample of older men. While item transformations and sample trimming based on a consistency test did improve the quality of the replication in this older, non-student sample, both methods have serious drawbacks. The five-factor solution appeared in the student sample without sample trimming or data transformation. Additionally, in both student and nonstudent samples, oblique rotation resulted in inter-factor correlations relevant to more general issues in the study of trait structure. We conclude that the 100 Unipolar Markers may be unsuitable for use in older populations or with nonstudent samples.