An analysis of gender DIF of the Gambling Behavior Scale for Adolescents (GBS-A) by using IRT models

In the assessment of adolescent gambling one central issue regards the measurement invariance across gender of the screens used to measure pathological behavior. Indeed, there is a general lack of evidence attesting their equivalent functioning in male and female adolescents. By employing Item Response Theory analyses, the present study was aimed at testing gender invariance of the Gambling Behavior Scale for Adolescents (GBS-A; Primi, Donati, & Chiesi, 2015), an instrument recently developed in order to measure the severity of Gambling Disorder (GD) as conceptualized in the last edition of the DSM (DSM-5; American Psychiatric Association, 2013). The GBS-A was administered to 1723 (56% males) 11- to 23-year-old students attending middle and secondary schools in Italy. Firstly, a confirmatory factor analysis demonstrated the unidimensionality of the scale in boys and girls. Then, IRT analyses for polytomous, ordered response category were applied. Discrimination and severity parameters indicated that all items were able to distinguish adolescents with different levels of GD and adequately covered the continuum of the latent trait. Subsequently, the equivalence of the items across gender was assessed by analyzing Differential Item Functioning, revealing that the GBS-A was gender invariant. Future studies might confirm the present findings by testing the GBS-A gender invariance with male and female adolescent gamblers recruited in clinical contexts. Overall, findings provided evidence that the GBS-A can be used to assess gambling problems in male as well as female adolescents.