TEAM TRAINING AND TRANSFER C OGNITIVE T ECHNOLOGY  V OLUME 12  I SSUE 2  F ALL 2007 17        Teams are used extensively in the workplace, and schools have been including teams in classroom learning. However, most theory-based learning research has focused on individuals. This research examined team-level knowledge acquisition and performance in two complex military tasks: An Unmanned Air Vehicle simulation and a peacekeeping simulation. Training occurred in co-located or distributed contexts, and testing occurred in the same or opposite context. Distributed performance was greater than for co-located teams. At initial transfer, all teams showed performance decrements but performance subsequently improved. Results for communication conditions (push-to-talk, no push-to-talk) revealed context effects at test, where using push-to-talk produced a greater decrement in performance. Improvements in both training and learning theory are suggested by identifying several variables that affect team performance. Key Terms: learning, training, transfer, team cognition, context In most real world situations, students learn in one context (school) and then are expected to apply their knowledge to a different context (job). Although transfer of learning has been studied extensively (for reviews see Lee, 1998; Singley & Anderson, 1989), most research has focused upon individual learning and transfer. However, most students are expected to learn in groups or teams, and often are expected to perform in teams in the workplace at a later time. Group learning has been studied at a social level, but team training and cognition has only recently become a focus for researchers (Cooke, Salas, Cannon-Bowers & Stout, 2000). Distributed learning has been available for some time with media such as videotape and correspondence courses, but increased use of the internet and the use of other new technologies have made distributed training of individuals and teams more readily available. Distributed team training can be a cost effective means of instruction; however, scholars have reported mixed results regarding team performance in distributed learning conditions (e.g., Hahn; 1990). By combining distributed learning and team training ideas and studying team learning and transfer in co-located or distributed contexts, a greater understanding can evolve regarding learning within collaborative situations and the role that context plays in that learning. Theory of Transfer Related to Context Change The predominant transfer theory derives from Anderson’s learning theory (Anderson, 1982, 1983). His theory predicts that transfer will occur when overlapping productions between two situations exist (Anderson, 1982, 1983, 1993). A production rule takes the form of IF (condition) THEN (action). For example, one might say, “If it is raining, then take an umbrella.” One can imagine that this particular production would be true whether one is in Seattle or in Tokyo. However, note that this might be different if a production were specifically tied to a certain context. For example, after an individual learns how to drive an automatic car, the same productions should fire whenever an individual is driving any automatic car. If, however, a person learns how to drive an automatic car but is then required to drive a manual (stick shift) car, then the productions do not overlap exactly and some transfer will not occur. Thus, when the context (driving a car) is held constant but the task is changed slightly (automatic to manual), a decrement in performance usually results (Singley & Anderson, 1989). One can imagine that if the context is vastly different then the condition portion of the productions would not overlap and performance should be greatly disrupted. Indeed, in studies of memory or state dependent learning, one can find up to 47% better recall of items when recall is in the same context as learning (Baddeley & Godden, 1980). Similar results have been found for contexts as obvious as drug induced states (Davies & Thompson, 1988; Eich, 1980, 1989) or as subtle as mood effects (Bower, 1983; Ellis, Adrienne Y. Lee New Mexico State University Gary D. Bond Winston-Salem State University Pamela S. Scarbrough New Mexico State University Douglas J. Gillan New Mexico State University Nancy Cooke Adrienne Y. Lee, Ph.D. New Mexico State University Gary D. Bond, Ph.D. Winston-Salem State University Pamela S. Scarbrough, Ph.D. New Mexico State University Douglas J. Gillan, Ph.D. New Mexico State University Nancy Cooke, Ph.D. Arizona State University