A new measure of group decision-making efficiency

It has widely been accepted that aggregating group-level decisions is superior to individual decisions. As compared to individuals, groups tend to show a decision advantage in their response accuracy. However, there has been a lack of research exploring whether group decisions are more efficient than individual decisions with a faster information-processing speed. To investigate the relationship between accuracy and response time (RT) in group decision-making, we applied systems' factorial technology, developed by Townsend and Nozawa (Journal of Mathematical Psychology 39, 321-359, 1995) and regarded as a theory-driven methodology, to study the information-processing properties. More specifically, we measured the workload capacity C_AND(t), which only considers the correct responses, and the assessment function of capacity A_AND(t), which considers the speed-accuracy trade-off, to make a strong inference about the system-level processing efficiency. A two-interval, forced-choice oddball detection task, where participants had to detect which interval contains an odd target, was conducted in Experiment 1. Then, in Experiment 2, a yes/no Gabor detection task was adopted, where participants had to detect the presence of a Gabor patch. Our results replicated previous findings using the accuracy-based measure: Group detection sensitivity was better than the detection sensitivity of the best individual, especially when the two individuals had similar detection sensitivities. On the other hand, both workload capacity measures, C_AND(t) and A_AND(t), showed evidence of supercapacity processing, thus suggesting a collective benefit. The ordered relationship between accuracy-based and RT-based collective benefit was limited to the A_AND(t) of the correct and fast responses, which may help uncover the processing mechanism behind collective benefits. Our results suggested that A_AND(t), which combines both accuracy and RT into inferences, can be regarded as a novel and diagnostic tool for studying the group decision-making process.