SMARTPHONE ADDICTION INVENTORY: DEVELOPMENT AND VALIDATION OF A REDUCED VERSION OF THE INSTRUMENT

Abstract

The objective of this study was to develop and validate a shortened version of the Smartphone Addiction Inventory (SPAI-BR), called SPAI-SV, for screening smartphone addiction in students. This is a methodological, epidemiological, and web survey study conducted with 1,998 students (1,321 from technical education and 677 from higher education). Data were collected exclusively via Google Forms between March and December 2021, using the original 26-item SPAI-BR. The shortened version of the SPAI-BR (SPAI-SV) ​​was developed using Exploratory Factor Analysis (EFA) applying tetrachoric brightness and oblique turns. The estimated factor structure obtained was completed by Confirmatory Factor Analysis (CFA) using Weighted Least Squares Mean and Variance Adjusted (WLSMV). The model validity was verified by cross-validation, dividing the sample into 70% for training and 30% for testing. Reliability (internal consistency) was assessed using the Kuder-Richardson coefficient (KR-20) and reproducibility using the Kappa coefficient and the intraclass brightness coefficient (ICC). The cutoff points for dependence classification were defined by the Receiver Operating Characteristic (ROC) curve. The EFA resulted in a final version of the SPAI-SV composed of 14 items distributed across three factors: Sleep Problems, Compulsion and Withdrawal, and Time Management and Daily Activities. The CFA confirmed the three-factor structure, presenting adequate robust fit indices, and was validated through cross-validation. SPAI-SV scores range from 0 to 14 points. A cutoff point of 5 or more points was identified as a possible dependency, with an Area Under the Curve (AUC) of 0.9768, sensitivity of 0.8781, and specificity of 0.9523, resulting in an accuracy of 91.79%. The instrument demonstrated high internal consistency (overall KR-20 = 0.829; 0.711, 0.783, and 0.701 for the three factors) and good reproducibility (Kappa = 0.663, indicating substantial agreement; ICC = 0.809, indicating excellent agreement). Thus, the validity, reliability, and reproducibility of the SPAI-SV were confirmed as a practical, robust, and efficient instrument for screening smartphone addiction in students. Its reduced version facilitates large-scale data collection and offers an objective and validated criterion for identifying individuals at risk, constituting a tool for psychologists, researchers, and educators in planning preventive interventions.