Internet severity and activities addiction questionnaire (ISAAQ): Psychometrics of item response theory and clustering of online activities

Background: Problematic usage of the internet (PUI) is an umbrella term, referring to a variety of maladaptive online behaviors linked to functional impairment. There is ongoing need for the development of instruments capturing not only PUI severity, but also the online activity types. The Internet Severity and Activities Questionnaire (ISAAQ), previously developed to address this need, required further refinement and validation.

Methods: Cross-sectional data was gathered in two separate samples (South Africa n = 3275, USA-UK n = 943) using the Internet Severity and Activities Addiction Questionnaire (ISAAQ). Item Response Theory (IRT) was used to examine the properties of the scale (Part A of the ISAAQ) and differential item functioning against demographic parameters. The severity scale of the ISAAQ was optimized by eliminating the poorest performing items using an iterative approach and examining validity metrics. Cluster analyses was used to examine internet activities and commonalities across samples (Part B of the ISAAQ).

Results: Optimization of ISAAQ using IRT yielded a refined 10-item version (ISAAQ-10), with less differential item functioning and a robust unidimensional factor structure. The ISAAQ-10 severity score correlated strongly with established measures of internet addiction (Compulsive Internet Use Scale [Person's r = 0.86] and the Internet Addiction Test-10 [r = 0.75]). Combined with gaming activity score it correlated moderately strongly with the established Internet Gaming Disorder Test (r = 0.65). Exploratory cluster analyses in both samples identified two groups, one of "low-PUI" [98.1-98.5%], and one of "high-PUI" [1.5-1.9%]. Multiple facets of internet activity appeared elevated in the high-PUI cluster.

Discussion: The ISAAQ-10 supersedes the earlier longer version of the ISAAQ, and provides a useful, psychometrically robust measure of PUI severity (Part A), and captures the extent of engagement in a wide gamut of online specific internet activities (Part B). ISAAQ-10 constitutes a valuable objective measurement tool for future studies.