An experiment in problem-solving was conducted in which one group of students used computer conferencing for continual public exchanges of ideas through asynchronous meetings, and another group used class discussion groups, while attempting to solve a legal problem in a course in consumer law. Student success rates for solving the assigned problem were similar. Based upon data collected, four parameters were identified as critical to integrating computer conferencing as an instructional strategy: (1) precise problem definition; (2) intensive preliminary training; (3) pacing student inputs/outputs; and (4) redistributing class time. The role of the classroom instructor was perceived as entirely different by members of each group. The mean scores and probabilities of student perceptions in each group are displayed in a table. (Author/ALF)