Early adolescence (typically aged 9-15 years) is a period of dramatic developmental change, and individual differences in temperament is likely to be an important predictor of the success with which individuals negotiate this period of life. Moreover, early adolescent temperament cannot be adequately captured by measures designed for other age groups. This study examined the empirical validity of the proposed temperament factors of the Early Adolescent Temperament Questionnaire-Revised (EATQ-R) in a large representative sample of 2,453 early adolescents aged between 10 and 12 years of age, and compared it with models that include cross-loadings between items and first-order factors, as well as first- and second-order factors. Furthermore, the reproducibility of the factor structure established by using a cross validation approach. Adding cross-loadings to the EATQ-R fit the data substantially better, resulting in an overall good fit that the original EATQ-R model did not achieve. However, the conceptual interpretation of the first- and second-order factor structures were not substantially altered even with this addition of cross-loadings. Future research should establish the construct validity of the first- and second-order factors as measured by this empirically based factor structure.
Keywords: cross-validation; early adolescence; factor structure; independent clusters factor pattern; temperament.