This paper reports on a two-year study conducted to evaluate cooperative learning or "group work" learning models in three courses at an Asian university. Data collection methods included a paired pre/post survey of learners, paired observations (early and late) of groups involved in assigned tasks, and two rounds of one-on-one interviews with a subset of learners. The study examined the effectiveness of the group work model on: student achievement of subject matter content; student changes in interpersonal and decision making skills; and student attitudes toward the model. The preponderance of significant changes from pre- to post-surveys on 164 matched questionnaires were negative, suggesting that the existing literature on cooperative or group learning may overstate the benefits such methods can provide. The study results indicated that: (1) group learning is no better than other instructional methods in overcoming problems of implementation; (2) the potential benefits of group learning may not be immediately recognized by students; (3) teachers wishing to employ group work methods must know what barriers their particular students face and take measures to help them overcome their deficiencies; and (4) research must be carefully designed if it is going to arrive at an accurate understanding of what transpires when students are given group learning tasks. (Contains 13 references.) (MAB)