企業經營環境日益複雜，公司許多重大決策，往往需要仰賴能夠集思廣益的團隊決策而非單打獨鬥的個人決策，然而團隊決策的決策成本，通常高於個人決策，因此，在何種情況下，團隊決策績效會優於個人決策？是一項值得探討的重要議題。本研究在於透過實驗室實驗，從受試者執行動態複雜任務的決策過程上，比較團隊決策與個人決策的差異。研究目的在於驗證：(1)團隊決策與個人決策在資訊處理過程上的差異，(2)團隊決策與個人決策在決策績效上的差異，(3)兩個潛在調節變項，包括「任務熟悉度」以及「決策目標的多寡」對於團隊與個人在決策績效上差異的影響。資料分析結果顯示：(1)動態複雜任務中，團隊決策與個人決策的資訊處理行為的確存在差異。相較於個人，團隧決策有較大量的資訊搜尋項目、所搜尋資訊有較高的比率是屬於決策攸關的歷史資料、需耗用較長的決策時間、決策錯誤的次數較少，並且在資源使用上，偏向積極而非保守。(2)就決策績效來看，團隧決策績效優於個人決策。(3)此種較佳的優勢，因任務熟悉與否而有所不同，在任務不熟悉的情形下，團隊的決策績效會優於個人，但是當任務變得熟悉後，團隊與個人決策績效間的差異便消失。(4)此種較佳的優勢，因決策目標多寡而有所不同，需要兼顧兩個目標時，團隊決策的績效會優於個人決策，但是當決策目標為單一時，團隊與個人決策績效間的差異則不明顯。

In the present study group performance was compared to individual performance through a laboratory experiment in which subjects were asked to complete a dynamically complex task. The study aimed to investigate (1) differences in information processing between individuals and groups, (2) individual and group differences on decision-making performance, and (3) the effects of two plausible moderator variables task familiarity and the number of goals, in the information processing model. Organizations increasingly highlight groups as an entity for decision-making in a highly dynamically complex business world. Relevant changes in management teams emphasize the collective over the individual and in job design teamwork that occupies an increasing proportion of work style to take the benefits of group interactions. However, decision-making by group often costs more than by individual. To understand when to use group instead of individual to make decisions is important. The results showed that (1) groups and individuals do differ in their information processing behaviors, with groups exhibiting more information search items, more historical-oriented data search, and longer decision-making time. (2) Groups generally performed better than individuals, however, (3) the difference was influenced by task familiarity. In unfamiliar tasks, groups revealed higher performance on decision making than individuals, while in familiar tasks, the difference disappeared. Meanwhile, (4) the number of task goals also showed impact on the performance between groups and individuals. Groups outperformed individuals when multiple (dual) goals were to be completed. The findings of this study display the following theoretical practical implications. First, the findings explain the reason why groups outperform individuals. Although past research has posited that it might due to the comparably superior information processing capability of groups, this study demonstrates this notion through an empirically sound research design. In this study a ”group” was composed of 4 individual members, each bearing their own unique experience, knowledge, and perspective attitude. This leads to more information search items and longer decision-discussion time than those of individuals, which broaden the scope and increase the depth of individual problem solving. The findings showed that group performance is better than individual performance in the following aspects: (1) attention procedure-groups have more and vast amount of information search items; (2) extraction procedure-groups process with more decision-relevant information; and (3) coding procedure-groups take longer decision time and makes less decision mistake. Moreover in the feedback procedure, groups were found to make more utilization of resources than individuals did. Regression analyses further confirmed the explanatory power of these information-processing variables to decision-making performance. Second, this study contributes to the knowledge body of decision making behavior in the perspective that group decision is an intervention to dynamically complex problems. Literature has documented that when encountered with dynamic complex problems, most subjects are unable to produce ideal solutions. Biased decision and low performance are commonly seen in dynamically complex tasks, especially in situations where specific structures and performance goals are imposed. Some scholars attempted to provide intervention to the problems by introducing decision-making facilitators such task feedback map, more structural information feedback, or suggestions for direction of decision-making. However, these interventions may not be of use in the real world, since decision-makers are often demanded to make immediate decisions or decisions under ambiguous situations. Our findings indicate that when subjects were dealing with unfamiliar tasks, without any facilitating tool aid, the mean performance by group is better than that by individuals. About 60 percent of the group medians are higher than individual medians. Only 20 percent of the individual medians are higher than those of the groups. In regard to the questioning of when individual decision making is appropriate, this study supports the notion that when environment becomes stable predictable or organizations are familiar with the changing tasks, a more simple decision process, for example, an individual decision-making or organization routines, can be appropriate. Our findings show that when strategic information was provided to subjects, there existed no difference between group and individual performance. Both individuals and groups were able to reasonably distribute resources in order for the simulated organization to grow. Third, this study identifies the importance of the moderator to group individual decision making performance, namely, the number of decision goals. Our results indicated that groups outperformed individuals in the situation where two decision goals were implanted, but not when there was only one single goal. This is probably because multiple goals increase the complexity of task, and that more capacity is needed for information processing, therefore group decision appears to be more effective in this condition. Fourth, business decisions as well as strategic decisions are usually dynamically complex in nature. Our findings yield the following implications for business management in that (1) group decisions are beneficial to dynamic complex problems; (2) group performance is not necessarily better than individual performance. Groups outperformed individuals in the aspects that they gather more information, pay more attention to historical information, and have longer discussion time. If one expects to enhance performance by means of group decision, the above circumstances have to be considered in favor of group decision before it is adopted. (3) Groups have a higher resource-utilization rate, indicating groups are more likely to make high-risk decisions. High resource-utilization promotes performance only when resource is reasonably distributed. One should pay cautiously attention to the risk of group decision in this respect. (4) As was revealed in the analysis of moderator, group decision does not apply to all situations. To save costs, group decision is a better practice only when unfamiliar tasks or multiple goals are demanded.