| Abstract: |
To understand youth character development, assessments must consider both the fluctuations of an individual's character across time and place (within-person variations), as well as the changes in their character relative to other individuals in their context (between-person differences). Although measures of character exist for between-person research, there is a need for context-specific, valid, and reliable measures of character that can capture both within- and between-person variations. In the present study, we used multilevel factor analysis (MFA) to establish the validity and reliability of a novel measure of character attributes--the Character Attributes Scale (CAS)--using an intensive longitudinal data set from the Compassion International Study of Positive Youth Development in Uganda. Participants (N = 200, 49.5% female, M[subscript age] = 12.13, SD[subscript age] = 2.19) completed a weekly survey of character attributes across 10 to 15 measurement occasions (total observations = 2,991). The MFA findings supported a four-factor structure for the CAS, consisting of 13 items at both the within- and between-person levels. These factors encompass the attributes of transcendence (three items), generosity (three items), gratitude (three items), and purpose (four items). In addition, the findings demonstrated significant positive associations between these attributes and intentional self-regulation, serving as an indicator of convergent validity. Reliability for subscales at the within-person level varied from [omega][superscript w] = 0.590 to [omega][superscript w] = 0.702 and at the between-person level from [omega][superscript b] = 0.802 to [omega][superscript b] = 0.881. Implications for future research using the current scale to study character development--including capturing both within-person variations (intensive assessments/idiographic approach) and between-person differences (cross-sectional assessments/nomothetic, differential approaches)--are discussed. |