Measuring Online Prosocial Behaviors in Primary School Children: Psychometric Properties of the Online Civility Scale

Children are online at increasing frequency, for longer periods of time, and at younger ages than ever before. Schools in the USA are engaged in providing online citizenship education to students; however, there are few evaluations of how well programs work or tools to measure program effectiveness. The current study provides psychometric information on one tool, the eight-item online civility scale (OCS), administered as part of an evaluation of the Be Internet Awesome (BIA), digital citizenship curriculum. The OCS was administered securely online to children in grades 4th, 5th, and 6th, to evaluate whether the BIA program improved their online citizenship skills. Psychometric analysis of the OCS was conducted for 824 youth who responded to both the pretest and posttest measures. Item constructs were validated in structural equation models with self-reported online behaviors. The eight items formed a cohesive construct, explaining 60% of the variance in online civility. The item response theory models showed that the items covered a range of the underlying trait. Differential item functioning (DIF) was observed by gender. The difficulty parameter did not show DIF across time (p = 0.11), but the discrimination parameter did (p = 0.002). The OCS construct was significantly and strongly associated with each of the validating variables in the expected directions. The OCS demonstrated excellent psychometric properties and could be a useful evaluation tool in programs designed to promote prosocial online behaviors in children. The scale behaves differently by gender and may discriminate differently across time points after an intervention.