Computer-related Causal Attributions: The Role of Sociodemographic Factors

Abstract

Attribution, i.e. a causal explanation for an event or behavior in situations of success or failure, has so far received little attention in HCI research. In this study we investigated whether computer-related attribution styles are related to sociodemographic aspects like age, gender, perceived computer skills and daily computer use. In total 66 participants (17-75 years) were asked to handle solvable (situation of success) and unsolvable (situation of failure) computer tasks. After each task they rated the cause of their success or failure on four attributional dimensions: locus, stability, controllability and globality. By using cluster analysis participants were grouped into three clusters for situations of success and failure, respectively and assigned to one of the six main attribution styles developed in prior research [1]. Finally, we investigated whether attribution styles are related to sociodemographic aspects by calculating Kruskal-Wallis tests. Results showed significant differences between young and elderly persons, whereby elderly persons showed more self-confidence and realism in their attributions than young. We also found significant differences between persons that self-rated their computer skills as poor and advanced. People with advanced skills attribute more often in a resigned style in situations of failure, whereas persons with poor skills attribute more in a confident style. Concerning gender and daily computer use we did not find any significant differences in computer-related attribution styles.