EFFECTS OF TASK STRUCTURE ON GROUP PROBLEM SOLVING

Abstract

This thesis investigates the effect of problem structure on performance and behavioural variety in group problem solving. In addition, it examines the effects of problem solving strategy in group problem solving. <br /><br /> Previous researchers have focused their efforts on individual problem solving with minimal reference to groups. This is due to difficulties such as the presence of distributed information, the coordination of people and the large scale of work that typified group problems. Specifically, the effect of problem structure in group problems has been rarely studied due to the absence of an encompassing theory. <br /><br /> In this thesis, the effect of problem structure on group performance is studied using the fundamental characteristics of structure such as detour, redundancy, abstraction and degree of homogeneity. These characteristics were used in conjunction with existing problem solving theories (such as Information processing system, Gestalt approach and Lewin's lifespace approach) and Heider's balance theory to understand the effects of task structure on group performance and behavioural output. <br /><br /> Balance theory is introduced as a conceptual framework in which the problem solving process is viewed as a dynamic progression from cognitive imbalance towards a state of structural balance corresponding with the solution. This theoretical approach captures both incremental search processes and insight associated with cognitive restructuring, typical of existing problem solving approaches in the literature. It also allowed the development of unique measures for studying the effect of structure in group problem solving. <br /><br /> A Laboratory experiment was conducted using 153 undergraduate and 3 graduate students in groups of 4 subjects. The experiment examined the effect of task structure on groups' performance and behavioural variety. The stimulus used for the experiment was a categorization problem consisting of sixteen cards with two objects each shared equally among four participants. The objective was to form four groups of items with no cards left unused. The groups' performance data was collected and analyzed to verify the postulated hypotheses. <br /><br /> The results indicate that both increased problem structure complexity and the introduction of a restructuring dimension in the problem structure were associated with reduced performance and increased behavioural variety. With respect to problem solving strategy, early discussion in problem solving was associated with better performance and less behavioural variety. Finally, the results support the premise that group problem solving processes tend to be in the direction of attaining higher states of balance.