(1) Background: Insomnia is one of the most prevalent sleep disorders worldwide and a growing public health concern, yet the combined contribution of psychological and demographic factors remains understudied in Latin America. This study aimed to examine how perceived stress, generalized anxiety, and sociodemographic variables predict insomnia severity in Ecuadorian adults. (2) Methods: A community sample of 698 participants (31.7% men; M = 25.6 years, SD = 10.4) completed an online survey during nationwide electricity blackouts in 2024. Measures included the Insomnia Severity Index (ISI), the Perceived Stress Scale (PSS-10), the Generalized Anxiety Disorder Scale (GAD-7), and a demographic questionnaire. Hierarchical linear regressions were conducted to evaluate the relative contribution of predictors. (3) Results: Demographic variables (age, sex, employment status, and socioeconomic level) were not significant predictors of insomnia severity (R² = 0.007). Perceived stress strongly predicted insomnia (ΔR² = 0.207, p < 0.001), and generalized anxiety added further explanatory power (ΔR² = 0.074, p < 0.001), with both factors contributing independently (β ≈ 1.96 and β ≈ 2.67). Interaction effects with socioeconomic status were non-significant. (4) Conclusions: Psychological factors explained nearly one-third of the variance in insomnia severity, underscoring the importance of stress and anxiety as core predictors and supporting the need for integrated psychological interventions to improve sleep health in Ecuador.

Executive functions (EFs) are vital for cognitive self-regulation and goal pursuit during adolescence, yet validated rating scales remain scarce in Latin America. This study evaluated the psychometric properties, item-level operating characteristics, and nomological network of the Amsterdam Executive Function Inventory (AEFI) in an Ecuadorian adolescent sample (N = 830; aged 12–18 years; M = 14.30, SD = 1.75; 52.77% female). Employing a split-sample derivation and validation approach with ordinal diagonally weighted least squares estimation, the original 13-item structure yielded poor fit. Following item refinement, a revised 10-item, three-factor model across attention, self-control, and planning demonstrated robust cross-validated structural fit (robust CFI = 0.970, robust TLI = 0.958, robust RMSEA = 0.048, SRMR = 0.043) and scalar measurement invariance across early (12–14 years) and late (15–18 years) adolescence. Dimension-specific graded response modelling confirmed adequate item functioning, with measurement precision concentrated across average latent trait levels (θ = −2.0 to 2.0). Nomological network analyses corroborated construct validity via strong associations with self-reported behavioural dysregulation and affective disruption on the SENA and PAI-A, alongside expected cross-informant divergence across family and school reports and limited discriminant separation from somatic complaints. These findings establish the revised 10-item AEFI as a promising, structurally invariant, and psychometrically adequate brief self-report measure for developmental and educational research in Ecuadorian youth.