The Deleterious Effects of Non-Response in Team-Level Research: What Every is Researcher Should Know to Avoid Bias

Abstract

Teams are an important part of the information systems (IS) field, in both practice and research. In practice, IS teams are the norm in software development. IS researchers have studied various aspects of IS teams, such as productivity and composition, among many others. In studying teams, researchers make the assumption that, in most cases, the aggregated responses from individual team members provide the ‘team-level’ data needed to examine the phenomenon of interest. Bias due to non-response is acknowledged, but rarely controlled or explicitly considered in the analysis. This study examines the effects that individual within-team non-response and the factors that condition its effect have in biasing team-level research. In our examination, a list of maxims to be considered or followed by researchers is produced. The explicit consideration of non-response bias in team-level research should help strengthen the research which, in turn, will help the IS field better utilize teams.

• non-response bias
• multi-level issues
• non sampling errors
• team-level
• group level