Factors Associated with Digital Tool Use and Perceived Learning Impact Among Secondary Education Students: Evidence from an Ecuadorian School Sample

This study examines the sociodemographic, technological, and pedagogical factors associated with the frequency of digital tool use and students’ perceived learning impact in secondary education. A quantitative, non-experimental, cross-sectional design was used with a non-probability convenience sample of 541 students enrolled in upper General Basic Education and the Unified General Baccalaureate in Ecuador. The statistical strategy was re-specified to address sparse response categories, non-proportional odds, multicollinearity, and marked institutional concentration. For frequency of use, ‘Never’ and ‘Occasionally’ were combined and a partial proportional-odds model was estimated. For perceived impact, the proportional-odds assumption was rejected globally, so two cumulative logits were estimated using Firth bias reduction. The two outcomes were positively associated (Spearman rho = 0.281, p < 0.001). Compared with students relying on a mobile phone only, those with a personal computer had higher odds of more frequent use (OR = 1.77, 95% CI 1.19–2.62) and of reporting significant perceived improvement (OR = 1.77, 95% CI 1.12–2.79). Reparameterisation around frequent reference categories showed that perceived teacher adaptation followed a non-monotonic pattern rather than a robust inverse gradient. The positive age association disappeared after accounting for its near collinearity with school year. Because the sample was convenience-based and 84.3% of participants attended one institution, the findings are interpreted as context-specific associations rather than population estimates.