L’objectif de cette étude est d’évaluer les propriétés psychométriques de la version française d’un questionnaire mesurant l’addiction à Internet, l’Online Cognition Scale, chez des étudiants français.

Un échantillon de 146 étudiants âgés de 18 et 38 ans (M=20,96 ; DS=2,65) a complété à deux reprises, à trois semaines d’intervalle, l’Online Cognition Scale, ainsi qu’une autre échelle mesurant l’addiction à Internet, et une échelle évaluant les symptômes dépressifs.

Une analyse factorielle de l’Online Cognition Scale a montré qu’une solution à cinq facteurs apparaissait comme plus appropriée que la solution factorielle originelle. Nos résultats mettent en avant une très bonne consistance interne de l’échelle ainsi qu’une corrélation inter-items satisfaisante (α=0,89 et 0,92 ; MIC=0,24 et 0,32). Les validités concordante, convergente et de construit étaient satisfaisantes.

La version française de l’Online Cognition Scale présente de bonnes propriétés psychométriques et apparaît comme un outil multidimensionnel adapté à l’évaluation d’une utilisation problématique d’Internet.

Internet is an essential tool for millions of people around the world, including in France where more than fifty million people use the Internet. Considering the increase of Internet use and the related consequences, many authors agree that excessive or improper use of this tool may lead to Internet addiction, a disorder which has been receiving increasing attention. Multiple theoretical frameworks have been used to define and evaluate this phenomenon, in particular substance dependence and pathological gambling. We chose to examine the Online Cognition Scale (OCS) which is a self-report questionnaire assessing Internet addiction based on a cognitive and behavioral model, with a multi- dimensional structure including four dimensions: diminished impulse control, loneliness/depression, social comfort, and distraction. While the psychometric properties of the OCS have been explored using English, Croatian, Turkish and Chinese versions, to date, the properties of the French version have yet to be examined. The aim of the present study was to evaluate the psychometric properties of the French version of the OCS in a student population.

A sample of 146 French university students aged between 18 and 38 years old (M=20.96 ; SD=2.65) completed the French versions of the OCS as well as another scale which measures problematic Internet use, and one which evaluates depressive symptoms. Participants completed the survey twice within an interval of 3 weeks.

Confirmatory factorial analyses revealed that the 4-factor solution did not demonstrate satisfactory fit, however, the 5-factor model was found to better fit the data. Our results revealed that the scale scores demonstrated excellent internal consistency and a satisfying inter-item correlation (α=0.89 and 0,92, MIC=0.24 and 0.32). The OCS was significantly correlated with the problematic Internet use at both time points (r=.73 and r=.74, P<.01), with time spent online (r=.34 and r=.45, P<.01) and with depressive symptoms (r=.31 and r=.28, P<.01).

The aim of this study was to evaluate the psychometric properties of the French version of the OCS in a sample of university students, assessing reliability and validity. Factors analysis revealed that a 5-factor solution including 27 items may be the most appropriate. Overall the scale revealed good psychometric properties. The internal consistency of the scale was consistent with previous validating results, which reported Cronbach's alpha ranged between .91 and .94. While the inter item correlations were satisfying, they were lower those previously found. The test-retest reliability of the OCS was good and satisfactory for all the factors, except for the Preoccupation and Withdrawal which revealed low values. These results are also lower than those found in other studies. A significant relationship was found between OCS scores and levels of depression as well as time spent online.

The French version of the OCS has satisfying psychometric properties and appears to be a useful multidimensional tool for the evaluation of Internet addiction.