Psychometric validation and cross-language invariance of the Arabic INCOM short form using PLS-SEM

Social comparison is a fundamental mechanism of social-cognitive functioning, yet validated instruments for assessing social comparison orientation in Arabic remain limited. This study presents a comprehensive, multistage validation of an Arabic version of the Iowa-Netherlands Comparison Orientation Measure (INCOM) using Partial Least Squares Structural Equation Modeling (PLS-SEM).

Study 1 employed a rigorous three-stage validation procedure: Stage 1 estimated a hierarchical measurement model specifying Social Comparison as a reflective higher-order construct comprising Ability and Opinion dimensions; Stage 2 addressed potential wording-related method effects by introducing positively rephrased alternatives for the two negatively worded items and re-estimating the model at the indicator level to reduce artifactual variance; and Stage 3 applied a prespecified retention criterion (outer loadings ≥0.70) to optimize reliability, convergent validity, and parsimony. Measurement quality was evaluated using indicator reliability, average variance extracted, heterotrait-monotrait ratio, and the Fornell-Larcker criterion, while predictive relevance was examined using PLSpredict. Study 2 employed a cross-sectional, between-groups design to examine cross-language measurement invariance between the Arabic INCOM short form and a matched English seven-item short form. Participants were assigned to one of two language versions (n = 251 per group), and invariance was evaluated using the three-step MICOM procedure based on permutation testing (5,000 permutations). Additional psychometric evidence was obtained through testretest reliability and associations with theoretically related external variables.

Findings supported a coherent seven-item short form (Ability: four items; Opinion: three items) with satisfactory psychometric properties. Study 2 examined cross-language measurement invariance between the Arabic short form and a partially modified English seven-item short form in two independent groups of university students. The Measurement Invariance of Composite Models procedure supported full measurement invariance, indicating configural, compositional, and scalar equivalence. Furthermore, evidence based on relations to other variables supported the scale through theory-consistent correlational associations with related constructs (neuroticism and social anxiety) and weak associations with theoretically more distal constructs (perceived social support and need for cognition).

Our findings suggest that the Arabic INCOM short form demonstrates satisfactory psychometric properties and cross-language measurement invariance, supporting its use for assessing social comparison orientation among Arabic-speaking university students.