Predicting Group Decision-Making with a Computerized Text Analysis Tool

Abstract

Decision-making tasks are often performed in groups because it is believed that a discussion of the topic from varying individual perspectives will lead to a better decision. However, researchers find group members often spend the majority of their discussion time focused on information members have in common rather than on unique pieces of information each group member holds (Stasser & Titus, 2003). Determining which groups are most likely to fall into this trap would benefit team leaders. A computerized text program, the Linguistic Inquiry Word Count (LIWC), offers one technological tool to meet this need. Significant correlations between LIWC word categories and time groups spent discussing common information, discussing unique information, and making the decision were found among 104 students who assessed candidates for student government president in face-to-face groups of two or four. The linguistic profile of groups following this unproductive pattern is identified and suggestions for further research are discussed.