Longitudinal Measurement Invariance of CCAPS-34 Scores with a Large University Sample

Abstract

Objective Counseling Center Assessment of Psychological Symptoms (CCAPS-34) scores were studied for longitudinal bias-free construct evidence. Method A sample of 4,696 university students referred to a university counseling center were assessed twice for evidence of longitudinal measurement invariance. Results Adequate or marginal longitudinal measurement invariance (LMI) of all subscales except for Eating Concerns was confirmed. Coefficients alpha and omega suggested that the internal consistencies for both initial and subsequent administrations were largely adequate for screening level test scores (≥ .80), except for the Hostility and Alcohol Use subscales (≥ .73). Second session effect sizes (Cohen’s d) were small but detectable and statistically significant for most subscales (-0.02–.17). Conclusions Counselors can use most CCAPS-34 subscale scores for tracking psychological symptoms and constructs over time with confidence. The Generalized Anxiety, Social Anxiety, Academic Concerns, and Alcohol Use subscales can be used with great confidence. Some confidence is warranted from the LMI evidence when using the Depression and Hostility subscales, but caution in warranted when using the Eating Concerns subscale.