Review by: Brahm, Taiga (2007-06-07)

The concept of efficacy has a long tradition in the learning sciences. Efficacy as the belief in one's ability to accomplish a particular behavior (Bandura, 1997) can be applied to a number of different context. The article reviewed focuses on collective efficacy in virtual teams which has not been studied yet. The main research question is how efficacy operates in a virtual team. The authors want to further explore prerequisites and consequences of virtual team efficacy. In their study of the literature, they found close relations between the concepts of group potency and collective efficacy. However, these two concepts should not be used interchangeably since group potency represents "the belief in the general effectiveness of the group across multiple situations" (p211) while collective efficacy is defined as the domain-specific belief in the group's competency, i. e. computer collective efficacy would be the belief that the group is able to work with computers. In particular, the authors developed the concept of virtual team efficacy, and defined it as "the belief in a team's ability to use communication technology to coordinate their activities across time and space" (p211). After defining these constructs, a research model to explain the antecedents and consequences of virtual team efficacy was developed. Besides computer collective efficacy and group potency, this model also included effort and communication level as mediating variables to explain the effect of virtual team efficacy on group outcomes and on actual performance. Prior research has identified the importance of mediators in the efficacy-outcomes relationship. Based on the empirical and theoretical literature, the following four hypotheses were developed:

• 1.) Group potency positively affects virtual team efficacy (H1).
• 2.) Computer collective efficacy positively affects virtual team efficacy (H2).
• 3.) Virtual team efficacy positively affects on group outcomes, mediated through its influence on team effort (H3).
• 4.) Virtual team efficacy positively affects actual performance, mediated through its effect on the communication level (H4).

The study aimed at developing and validating instruments for the different constructs. The development of the measures was based on literature with the scale for virtual team efficacy being created as a new scale. Effort und group outcomes were operationalized as perceptual variables while the communication level and actual performance were assessed with objective measures (number of discussion postings and grade for the group work). The scales developed were tested with data from several virtual team projects. The teams consisted of students from universities in the United States, Hong Kong, and Great Britain. In total, there were 318 usable responses from 52 project teams. The empirical study followed the common procedure of item development, exploratory factor analysis, and confirmatory factor analysis. After modifying the number of items during the confirmatory factor analysis, the model provided a very good model fit with good indicators for convergent and discriminant validity. After this first model test, the data were aggregated on the group level to test the hypotheses. For this, partial least squares (PLS) estimation was used since this method is more reliable for small samples than the covariance-based maximum likelihood method. This analysis provided support of the hypotheses H1, H2, and H3, however, H4 was only partly supported: Virtual team efficacy indeed had a positive effect on actual performance, however, the communication level did not fully mediate this relationship.

In sum, the research article provided new insights for virtual team research by developing and testing an instrument for virtual team efficacy. The assessment of the proposed research model provides further evidence of some of the proposed relations and puts virtual team efficacy into a comprehensive research framework. Limitations of the study include the usage of student teams instead of real-life virtual teams. Additionally, some other constructs may be of further interest for the relationship between efficacy beliefs and virtual team success (e. g. trust, leadership, conflict management). The study did not control for extragenous variables such as task type, group size, and project durations leading to more confidence in the generalizability of the findings.