Adicção em Internet: propriedades psicométricas do Internet Addiction Test e prevalência em estudantes

Abstract

Introduction: The advent of the internet has provided numerous benefits, but its inappropriate and excessive use can lead to the development of a dependency disorder called Internet Addiction (AI). Objective: To evaluate the psychometric properties of the instrument Internet Addiction Test (IAT) and to estimate the prevalence of Internet Addiction (AI) among students. Two studies were carried out with population composed of high school and higher education students from public and private educational institutions in the municipality of Montes Claros, MG, Brazil. Study 1: Methodological study that evaluated the validity and reliability of the IAT in the population studied. Validity was evaluated through the analysis of psychometric sensitivity, construct validity (factorial, convergent, discriminant and hypothesis testing) and criterion validity (concurrent); stability of the adjusted model was evaluated in independent subsamples through multigroup analysis; reliability was evaluated by means of internal consistency and reproducibility. A total of 2519 randomly selected students participated in the study, being 29.6% and 70.4% from high school and higher education institutions, respectively. All IAT items showed normal distribution indicator values (asymmetry coefficients <3 and kurtosis <7) and, therefore, adequate psychometric sensitivity. A three-factor model was adjusted: "Emotional and cognitive preoccupation with the internet" (Factor 1), "Time management problems" (Factor 2) and "Performance problems" (Factor 3), which obtained satisfactory indexes of adequacy and stable structure in the independent subsamples. Average Variance Extracted values obtained in the analysis of convergent validity were equal to 0.32, 0.41 and 0.45 and Composite Reliability values equal to 0.84, 0.81 and 0.71, for Factor 1, Factor 2 and Factor 3, respectively. As for the hypothesis test, the IAT was able to distinguish different levels (p <0.05) of AI among students according to gender, age, socioeconomic class, frequent use of social networks, online games, videos/movies and music. Concurrent validity was shown with positive and significant correlations (p <0.05) between the IAT score and self-perception of AI (r = 0.427), time spent online on working days (r = 0.427) and weekends (r = 0.424). Internal consistency was adequate (α = 0.906) as well as reproducibility (Kappa = 0.73 and Intraclass Correlation Coefficient = 0.90). The IAT version, translated into Portuguese, presented satisfactory psychometric properties and can be considered valid and reliable to evaluate AI in Brazilian high school and university students. Study 2: A cross-sectional, population-based study that estimated the prevalence of AI and associated factors in the population studied. The AI was evaluated through the Internet Addiction Test (IAT) and the factors analyzed were: demographic, socioeconomic, lifestyle and internet usage profile characteristics. Descriptive analysis of the variables (frequency distribution, center measurements and variability) was performed. Prevalences of AI were estimated, with a 95% confidence interval (95% CI). In order to identify the factors associated with Problems with Internet Addiction (IAP), Poisson regression model was used, with robust variance, which allowed the estimate of prevalence ratios (PR). Of the 2519 participants, 6.2% [CI95%: 4.3%, 8.8%] were considered to be internet addicts and 46.1% [CI95%: 42.2%, 50.0%] problematic users, totaling 52.3% [CI95%: 46.8%, 57.7%] with IAP. In high school, having an IAP was positively associated with: female gender (PR = 1.12), alcohol consumption (PR = 1.33), impaired sleep (PR = 1.24), frequent use of social networks (PR = 3.16) and connection time (RP = 1.03). Negative association with age over 16 years (PR = 0.90) and economic class C (PR = 0.73). In higher education, the following factors were positively associated with IAP: female gender (PR = 1.12), paid activity (PR = 1.14), alcohol consumption (PR = 1.24), impaired sleep (PR = 1.39), imbalanced diet (PR = 1.19), frequent use of social networks (RP = 2.57) and time connected (RP = 1.04). Negative association with age over 23 years (PR = 0.70), economic class C (RP = 0.89) and D / E (RP = 0.66). Over half the students presented problematic behavior regarding the use of internet, which was shown to be associated with demographic, socioeconomic, lifestyle and use of internet characteristics.