Browsing by Author " Soto Martínez, Marcela I."

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Comparative Analysis of Etiological models of Adolescent substance use
    (2025) Soto Martínez, Marcela I.; Vergés, Álvaro; Pontificia Universidad Católica de Chile. Escuela de Psicología
    Drug use is a mental health concern with serious physical and psychological consequences, which are especially critical during adolescence. In Chile, approximately 20% of adolescents report alcohol or marijuana use during the last year. For both substances, prevalence is higher among females (alcohol: 26%, cannabis: 21%) than males (alcohol: 20%, cannabis: 17%). Use increases with grade level—from 11% (alcohol) and 10% (cannabis) in eighth grade to 36% and 27%, respectively, by the final year of high school. Additionally, students attending private schools report higher alcohol use (30%) but lower cannabis use (14%) compared to those from other school types.To explain individual differences in substance use patterns, several mediational models have been proposed, focusing on the role of cognitive variables in the relationship between impulsivity and substance use. Two key models in this field are the Acquired Preparedness Model (Smith & Anderson, 2001) and the 3-Component Approach to Reinforcing Substances (3-CARS; Gullo et al., 2010).This study aimed to test and compare the APM and 3-CARS as etiological models of substance use behaviors in adolescence. To this end, three studies were conducted: (1) a systematic review of the literature on impulsivity, expectancies, and substance use in adolescents; (2) an evaluation of the psychometric properties of the Cannabis Refusal Self-Efficacy Scale; and (3) an empirical comparison of the APM and 3-CARS in a Chilean adolescent sample.The systematic review included 17 studies, 23 models, and 123 mediations, focused exclusively on alcohol and tobacco. Few studies used longitudinal designs (n = 4) or probabilistic sampling (n = 7). The most frequently examined impulsivity-related constructs and mediators were sensation seeking (n = 47), impulsivity (n = 31), reward drive (n = 18), positive social expectancies (n = 28), coping motives (n = 28), and refusal self-efficacy (n = 26). Of all mediations analyzed, only 37.6% were statistically significant (n = 50), whereas over half of the direct effects were significant.The validation of the Cannabis Refusal Self-Efficacy Scale yielded good model fit: χ²(74) = 245.662, p < .000, CFI = .974, TLI = .969, RMSEA = .047, SRMR = .029. Psychometric indicators were strong, with factor loadings ranging from .65 to .94, reliability coefficients from ω = .64 to .95, and temporal stability (ICC) between .78 and .88. As expected, the scale was significantly correlated with marijuana use frequency, problems, and reinforcement sensitivity. Scalar invariance was established across sex, age, and problematic use levels, enabling meaningful group comparisons.In the final empirical study, four models were estimated (2 theoretical models × 2 substances: alcohol and cannabis), all showing good fit. The 3-CARS models explained more variance in problematic cannabis use, and included a greater proportion of significant direct effects than the APM. Across all models, positive expectancies emerged as the only consistent significant mediator. The only model in which no significant indirect effects were found was the 3-CARS model predicting cannabis-related problems.This study is the first to compare the APM and 3-CARS models and to evaluate their predictive capacity for cannabis-related outcomes. Reward drive and sensation seeking—identified as significant mediators in the empirical analysis—were also the most frequently studied impulsivity constructs in the systematic review. Positive expectancies consistently emerged as a key mediator, supported by both theoretical and empirical evidence. Finally, theoretical and practical implications are discussed.