Teach Talk 

From a number of different quarters the call has come for science and engineering students to be educated in skills beyond the traditional analytical/technical/problem-solving abilities that have been the mainstay of an MIT education. This trend was, in part, the impetus behind the new Communication Initiative. But education in the so-called "soft skills" includes not only training in writing and public speaking; it also entails improving social and interpersonal communication skills such as leadership and the ability to work with others in teams.

The recognition of the importance of these kinds of skills in providing a complete education comes not only from outside agencies like the NSF and ABET, both of which have lobbied for their inclusion in science and engineering curricula, but it also comes from MIT faculty, alumni, and students. For example, in "The 1994 Senior Survey Report," prepared by Alberta Lipson, Norma McGavern, and the Educational Studies Working Group, 70 percent of the students surveyed thought the ability to work in teams was "very important" (although only 58 percent said that skill was improved "moderately" or "greatly" during their four years at MIT). And in the "Report to the Education Committee of the Academic Council," written just this past December, the authors, the Faculty Policy Committee, issued the challenge that "[MIT] will have to model an environment where its students . . . learn communication skills . . . and have the experience of interacting with all kinds of people." (p. 1.)

But teaching students to communicate with, listen to, and work well with one another isn’t simply a matter of putting them in groups and letting them go about their business. Like any skill, the ability to work effectively with others requires some basic knowledge of best practices (in this case of team dynamics, consensus building, decision making, etc.); a chance to practice the skills to be acquired; and feedback on the success of these efforts. It also requires that instructors think about how to structure teams, assignments, and class time so that the teamwork experience is optimized and the most learning that is possible occurs.

Fortunately, in recent years some excellent research has been done in using teamwork as a pedagogical tool, and several faculty members at MIT have run extremely successful experiments in using student teams. This Teach Talk and the one to follow will provide some guidelines on how best to work with student groups. Some of this information was presented in an IAP workshop "Teaching Teamwork Skills" that was part of the "Better Teaching @ MIT" series this past January. A similar workshop can be arranged for individual faculty through the Teaching and Learning Laboratory.

A distinction before beginning this discussion: All sorts of groups can be used successfully in the classroom. Students can be put together informally for one class to work on a problem or check homework. They can meet outside of class in study groups to prepare for a test. Here, however, I am going to focus on student teams that operate for longer periods of time (usually the entire semester), working on either several short assignments, on one large project, or both. This kind of experience--in which students are asked to interact with the same classmates over the course of several months--allows team members to watch the group process unfold, and, hopefully, permits trust and cohesiveness to build, which is at the foundation of all successful teamwork

What helps ensure that this goal will be achieved? Here is some of what we know about how to use teams productively in the classroom:

Assign the teaching staff the responsibility of creating student teams. If students are allowed to choose their own groups, it is only natural they are likely to team up with friends, housemates, or people living close by. While it may be helpful to have teammates who live near one another (one of the most frequent complaints students have about teamwork is the difficulty of arranging meetings outside of class), the obvious disadvantage of permitting students to select their own teams is that they won’t have the opportunity to work with people with whom they’re not familiar – a skill they will need once they enter the working world.

Decisions about team composition should be made according to several criteria. First, decide how large the teams should be. Four to five people is generally the number that is advised, but there may be good reasons to make groups larger or smaller. In making that decision, think about how much work you are requiring, the time frame in which it is to be done, and logistics (e.g., are there limitations on the physical space in which the team will have to operate?).

Second, determine how homogeneous or heterogeneous you want the teams to be. In fact, you need to decide what defines homogeneity or heterogeneity - differences in gender and ethnicity? intellectual interests? personality characteristics? Here, too, thinking about educational objectives, course requirements, and logistics will help you work out an effective strategy.

Provide some training in teamwork skills. Students don’t come with a built-in facility for working efficiently with others. (For that matter, most adults don’t either!) Again, students need to be taught the interpersonal communication and team building skills that will help to ensure smoothly functioning groups. And we have a responsibility – if we expect them to work together – to provide that training.

In a pilot project Professor Mary Boyce and I did to bring teamwork and active learning to "Mechanics and Materials II" (2.002), students were given a two-hour teamwork workshop during one of their regularly scheduled laboratory classes. The workshop covered such topics as setting norms, using a facilitator, listening, and giving and receiving feedback; it also gave students some guidance on how to deal with common problems teams often encounter. Students were given a manual that provided more information on these topics, as well as several readings on conflict resolution, decision making, and the interplay between personal characteristics and team participation. (The next Teach Talk will describe a comprehensive training program in teamwork skills that has been developed by Chemical Engineering Lecturer Bonnie Burrell as part of the Chemical Engineering "Projects Laboratory" (10.26) and "Processes Laboratory" (10.27).)

Coach the teaching staff in teamwork skills. If recitation instructors or teaching assistants are assigned to the class, they should also have some basic training in teamwork skills. In the 2.002 experiment, we had a two-hour training session for both laboratory instructors and TAs that covered some of the topics listed above, as well as particular problems the instructional staff might face in working with the student teams. We also presented the participants with scenarios based on three of those common dilemmas and asked the group to brainstorm possible solutions.

Give students time to work together in class. According to Management Professor Larry Michaelson, who has done extensive research on using student teams, "A key to effectively using learning groups is using practices that promote the development of group cohesiveness." ("What Every Faculty Developer Needs to Know about Learning Groups," To Improve the Academy, Volume 15, 1996.) One of the ways that group solidarity is fostered is simply by having group members do things together. But, as Michaelson points out, the "cost" of students getting together outside of class is often so high that the most common practice is for students to meet only to divide up the work at hand, and then to do those tasks independently. Giving students time to get together in class guards against that.

In a course taught last semester, "The Structure of Engineering Revolutions" (STS.185/6.972), Professors David Mindell and Charles Leiserson devoted 11 out of 25 classes to group work. The first two of those classes were in training on teamwork and collaborative writing. The next three focused the groups on very specific tasks as a way to break the ice and give team members an opportunity to learn about each other’s capabilities. Three more classes allowed for the groups to work on their semester projects, and the last three classes were devoted to group presentations.

According to Professor Mindell, "There’s no question that giving the groups time to meet in class improved their effectiveness. It also gave us the opportunity to oversee their work and help out where we could." However, Mindell did say that the drain on class time was a difficulty, and next year a discussion section would be added to the course for group meetings.

Structure assignments so students must work together. If assignments are devised so that students can simply divvy up the work and do it on their own, they are likely to do just that. (Although even assignments that permit this kind of individual effort will at least require students to coordinate their work just before the assignment is due in order to submit a cohesive package.) In the 2.002 course, Professor Boyce attempted to address this problem, in part, by creating assignments that required students to work together in the laboratory. Michaelson suggests the best assignments to foster group interaction "require students to make a decision with respect to a complex set of data" or to "make a concrete decision based on an analysis of a complex issue." For example, in assignments where students are asked to manipulate data in order to come up with a single decision or recommendation, team members must practice decision making and consensus building skills in order to complete the assignment satisfactorily.

Give feedback throughout the semester. When students have no way to gauge how well they are doing, groups can flounder. Providing immediate, unambiguous feedback helps to promote both team development and learning. Some controversy exists over whether or not feedback should be made public in order to allow comparisons with other groups. Michaelson maintains that since the "single most powerful force for the development of group cohesiveness is the presence of an outside influence that is perceived to be threatening to member goals and/or the well-being of the group," providing this kind of comparative performance data has a beneficial effect. ("Designing Effective Group Activities," To Improve the Academy, Volume 16, 1997.) However, that strategy may promote a kind of competitiveness in your class that is antithetical to a spirit of collaboration you may be trying to foster.

It is fairly clear, though, that if groups are to work on one large project throughout the semester, then the project needs to be broken into smaller assignments that have to be handed in throughout the semester, so students can get a sense of how the group is doing And, ideally, feedback should be on both the work accomplished and the group processes that allowed it to happen.

Next Teach Talk: the instructor’s role in the group process, providing mechanisms for reflection on group work, and grading.