Emotions and Performance in Virtual Worlds. An empirical study in the presence of missing data

In this work, we first investigate characteristics of virtual worlds and determine important situational variables concerning virtual world usage. Moreover, we develop a model which relates individual differences of virtual world users, namely emotional and cognitive abilities, experiences with virtual worlds as a child, and the level of cognitive absorption perceived during virtual world use, to the users’ individual performance in virtual worlds.We further test our model with observed data from 4,048 study participants. Our results suggest that cognitive ability, childhood media experience, and cognitive absorption influence multiple facets of emotional capabilities, which in turn have a varyingly strong effect on virtual world performance among different groups. Notably, in the present study, the effect of emotional capabilities on performance was stronger for users which prefer virtual worlds that have more emotional content and require more social and strategic skills, particularly related to human behavior. Interestingly, while cognitive ability was positively related to various emotional capabilities, no evidence for a direct path between cognitive ability to performance could be identified. Similarly, cognitive absorption positively affected emotion perception, yet did not influence performance directly. Our findings make the case for abandoning the traditional perspective on IS–which mainly relies on mere usage measures–and call for a more comprehensive understanding and clearer conceptualizations of human performance in psychometric studies. Additionally, our study treats missing data (an inherent property of the data underlying our study), links their presence to theoretical and practical issues, and discusses implications