Is Social Learning Always Helpful? Using Quantile Regression to Examine the Impact of Social Learning on Information Security Policy Compliance Behavior

Abstract

Social learning theory has attracted increasing attention in recent years in terms of its use to study information security policy non-compliance behavior. But previous results of studies in the field of information security have been rather heterogeneous. Various influencing factors have been considered within the framework of social learning theory. Previous studies quantitatively assess the effectiveness of social learning by relying on mean-based regression methods. In contrast, we intend to apply quantile regression to provide a new perspective on the subject. Therefore, we estimate the overall impact of social learning interventions and uncover how their impact differs among employees with different propensities (quantiles) for information security policy compliance behavior—an important finding for determining safety interventions for specific employee groups. Based on data collected in Germany, our results show significantly different effects in the analyzed quantile aspects of imitations and differential reinforcement.