A Multi-Level Approach to STEM Identity: Conceptualization and Measurement of Personal and Social Dimensions of STEM Identity in Early Adolescence

This study draws on a developmental social-psychological perspective of identity to inform the conceptualization and measurement of STEM identity in early adolescence. In doing so, we emphasize that STEM identity is both personally and socially constructed and we provide a way to operationalize these dimensions by adapting two well-established instruments – the Utrecht-Management of Identity Commitments Scale (U-MICS) and the Group Identification Scale (GIS) – to assess personal and social STEM identity. The sample included 233 Romanian 6th and 7th graders (Mage = 13.22, SDage = 0.57, 50.2% boys). Using a three-wave design, spaced 3 to 5 months apart, factor structure and longitudinal measurement invariance were examined. Using the multiple-indicators multiple-causes method, we tested differential item functioning across age and gender. For both scales, results supported theoretically consistent factor structures and high internal consistency. Item responding was not influenced by age or gender. The U-MICS for STEM exhibited full scalar invariance, and the GIS for STEM exhibited partial scalar invariance. These findings contribute to the study of personal and social identity in STEM contexts by providing psychometrically valid and developmentally grounded measures, which can be used in future longitudinal and intervention studies examining STEM identity from a developmental social-psychological perspective.