Early detection of problematic internet use among university students: psychometric validation of the PIUQ-9

University students extensively use the internet for academic, social, and recreational purposes, increasing the risk of developing problematic internet use (PIU). Although several instruments exist, there is a need for brief, psychometrically sound screening tools capable of identifying individuals at risk and detecting related conditions such as Internet Gaming Disorder (IGD) and Online Gambling Disorder (OGD).

This study aimed to examine the reliability and validity of the 9-item Problematic Internet Use Questionnaire (PIUQ-9) in a sample of first-year university students and to explore its screening performance for PIU and its potential utility for identifying increased risk of related conditions such as Internet Gaming Disorder (IGD) and Online Gambling Disorder (OGD).

A cross-sectional study was conducted with N = 273 university students [180 women (65.9%), 93 men (34.1%); mean age 19.5 years, SD 1.9]. Participants were recruited from university settings and completed a self-administered battery of validated instruments assessing PIU, addictive behaviors, impulsivity, emotion regulation, and general psychopathology. Construct validity was examined using confirmatory factor analysis. Internal consistency was assessed using McDonald's omega. Convergent validity was evaluated through Pearson correlation coefficients with related constructs. Test-retest reliability was assessed using intraclass correlation coefficients (ICC) and weighted kappa statistics. Receiver operating characteristic (ROC) curve analyses were conducted to determine empirically derived screening threshold for identifying high risk of PIU, IGD, and OGD.

Confirmatory factor analysis supported a unidimensional structure of the PIUQ-9. The scale demonstrated good internal consistency (ω = 0.82) and explained 38.5% of the total variance. Convergent validity was supported by moderate to strong correlations with related measures, including risk of internet/social network addiction (r = 0.666; 95%CI: 0.594-0.727), impulsivity (r = 0.369; 95%CI: 0.262-0.467), emotion regulation difficulties (r = 0.382; 95%CI: 0.276-0.479), and psychological distress (r = 0.312; 95%CI: 0.201-0.415) (all P < 0.001). Test-retest reliability was good (ICC = 0.753; 95%CI: 0.650-0.830), with item-level agreement ranging from moderate to substantial (weighted kappa = 0.435 [95%CI: 0.278-0.593] to 0.698 [95%CI: 0.593-0.803]). The area under the ROC curve was 0.824 (95% CI: 0.768-0.881). ROC analyses identified an empirically derived screening threshold of 9 for high risk of PIU, 11 for increased risk of IGD, and 13 for increased risk of OGD. These thresholds showed variable discriminatory performance, with lower specificity and positive predictive values for IGD and OGD.

The PIUQ-9 is a brief, reliable, and valid instrument for assessing PIU among university students. This study extends previous research by providing evidence of its accuracy in identifying related behavioral addictions, such as IGD and OGD, within a population highly vulnerable to PIU yet still underrepresented in the scientific literature. The findings support its utility as an efficient screening tool in both research and applied settings, facilitating early detection and targeted prevention strategies in young adult populations.