Integrated Psychometric Evaluation of the Self-Report SDQ Emotional Problems Scale Across Adolescence

Abstract: Despite the widespread use of the Emotional Problems subscale of the Strengths and Difficulties Questionnaire (SDQ-EP), integrated evidence on its longitudinal measurement properties remains limited. The present study used two longitudinal school-based samples of adolescent self-report data from Norway (Grades 8–10, N = 2,079; Grades 11–13, N = 3,102) to examine structural validity, longitudinal and gender measurement invariance, and item functioning within a unified measurement framework. Wave-specific confirmatory factor analyses provided qualified support for a broadly unidimensional structure, although model fit was weaker and residual associations were more pronounced in the younger cohort. Longitudinal partial scalar invariance was supported in both samples after freeing one threshold for a single item at one measurement occasion, and gender invariance was supported within the resulting longitudinal models. Graded response models demonstrated moderate-to-strong item discrimination and systematic differences in threshold location, with somatic complaints showing lower discrimination than cognitive-affective items. In the younger cohort, the SDQ-EP showed preliminary evidence of discriminant validity relative to a brief measure of affective distress, although the strong association between the two measures indicated considerable overlap. Overall, the findings provide qualified support for using the SDQ-EP in longitudinal and gender-based adolescent research.