Answers are grounded in this paper and cite their sources.
This study reassessed the validity of the AI Dependence Scale among 316 secondary school students (51.3% male; age 16–19 years, M = 17.30) from three educational directorates. Participants completed measures of AI dependence, reading comprehension beliefs, and AI-use stigma. Construct validity was examined using exploratory factor analysis (principal axis factoring with Promax rotation) and confirmatory factor analysis (DWLS estimation), with reliability assessed via Cronbach's α and McDonald's ω, and concurrent validity assessed via Pearson's correlations. A unidimensional structure was confirmed, accounting for 54.7% of variance with loadings of 0.654–0.782 (EFA) and 0.704–0.835 (CFA). Confirmatory factor analysis (DWLS estimation) supported this structure, demonstrating acceptable fit across most indices (CFI = 0.957, TLI = 0.945, NFI = 0.950, SRMR = 0.046). Although the RMSEA value was elevated (0.131, 90% CI [0.115, 0.148]), this inflation is statistically expected given the low degrees of freedom (df = 35) in this parsimonious unidimensional model. Reliability was excellent (α = 0.923, ω = 0.924). Concurrent validity was supported by significant correlations with stigma (r = 0.275, p < .001) and reading comprehension beliefs (r = 0.295, p < .001), whilst these constructs remained empirically distinct (r = 0.088, p > .05). The findings confirm the scale as a robust unidimensional instrument, challenging assumptions of multidimensionality and suggesting that dependence manifests as a generalised orientation during adolescence. The scale offers a parsimonious, reliable tool for investigating technology dependence, with implications for educational practice and future research.
No takes yet. Share an insight, caveat, or question.
Amer et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: