Short Version of the Social Networks Addiction Risk Questionnaire (CARS-R): Evidence From Network Psychometrics
Author(s)
Lindsey W. Vilca
Tomás Caycho‐Rodríguez
Jessica Aranda Turpo
Carla Dávila-Valencia
Emily Lupaca-Huarac
Date Issued
18 de octubre de 2025
Type
Article
Start Page
332941251389577
End Page
332941251389577
Abstract
Recent scientific literature shows evidence that social media addiction negatively impacts people's mental health. In this sense, network models offer a methodological framework that naturally allows the study of mental health disorders' complexity. Therefore, the study aims to study the psychometric properties of CARS-R from the perspective of psychometric networks. A sample of 1,719 adults of both sexes (38.3% men and 61.7% women) between the ages of 18 and 59 years (M = 21.9; SD = 6.13) was collected. The EGA analysis showed that the items form a single community composed of nine nodes with large network loading values (>.35). The UVA analysis showed that all items are relevant to the network. Evidence was also found on the reliability of the scale, where the items are stable and systematically organized into a single community 100% of the time. Furthermore, bootEGA analysis showed that CARS-R is invariant across sex, age, and hours of use. The nodes were also found to have positive relationships, and the centrality index identified nodes C2, C9, and C7 as the most influential in terms of strength. A new definition of the construct is proposed based on the results found. It is concluded that the results of this study show solid evidence of the functioning of CARS-R and allow us to propose a new definition of social media addiction.
Subjects