A proposed method for the study of predictors for detecting excessive use of technology: Problematic Internet and mobile phone usage in adolescents

This study uses an innovative statistical strategy to test the role of certain variables as predictors of problematic Internet and mobile phone usage among adolescents in Spain and in the United Kingdom . A paper-and-pencil questionnaire was used, with socio-demographics and patterns of technology usage as variables, and two tests were administered: the Problematic Internet Entertainment Use Scale for Adolescents (PIEUSA) and the Mobile Phone Problem Use Scale for Adolescents (MPPUSA). The overall sample size was 2228 high school students aged between 11 and 18 from Barcelona and London. PIEUSA and MPPUSA scores were transformed into normed scores, and both were then dichotomized according to three statistical criteria as cut-off points (i.e., median, 80th percentile, and extreme scores below the 25th percentile and above the 75th percentile) in order to establish the relationship between the variables above and the excessive use of the Internet or mobile phones, using a binary logistic regression. The results show that the best predictive model for both technologies includes socio-demographic variables as predictors of extreme scores for excessive Internet and mobile phone usage, with good sensitivity, specificity and classification accuracy, as well as a notable capacity for discrimination according to the receiver-operating characteristic curve. Implications of these findings are discussed.