3.5. FORMATION OF TEAMS
There are many attributes that need to be considered when forming a team. The work of Pelz and Andrews (1966b) has already indicated that effective teams are characterized by the support of members for one another's work, great respect for other members, and complementary skills, strategies, and approaches. Respect requires similarity in level of ability, as well as similarity in basic attitudes and values. Yet team members must have varied skills and specific attitudes that are different and complementary. For example, they may all be top-level scientists with similar values toward autonomy, yet differ in their disciplines and in their attitudes toward specific methods of data collection. Empirical support for this point has been obtained in laboratory studies (Triandis et al., 1965), indicating that creativity was highest when team members were similar in their abilities but different in their specific attitudes.
Research is increasingly becoming interdisciplinary and many of the research teams include members from different disciplines and cultures. That creates problems of coordination, but may also increase creativity. Including team members who are multicultural may increase creativity. As Leung and colleagues (2008) argue, multicultural experience can contribute to creativity because it enables people to (a) broaden the range of accessible ideas and concepts, (b) recognize that the same form can have different functions, (c) destabilize established associations, (d) increase their readiness to seek out ideas from diverse sources, and (e) build greater cognitive complexity.
Similarity in ability is important because it is undesirable for people to look at their co-workers as being intellectually too different. A person feels uncomfortable with co-workers who are either "dummies" or "geniuses," and thus cooperation will suffer. At the same time, complementarity is desirable for many personality attributes. For example, a person who would like to dominate a team can get along much better with people who like to be dominated than with others who also want to be in charge. A person who is talkative gets along better with those who like to listen. A diversity of viewpoints on how to approach a research project and different skills concerning its actual operation may result in a better research project (Janis, 1972). If members of the team are too different, cooperation will suffer; if they are too similar in perspective, groupthink may result (see Chapter 2).
What is the optimal team number? There is research suggesting that five is the best number for a discussion group. However, McComb, Green and Compton (2007) imply that there is no set number. Rather the level of flexibility found within a team affects staffing quality. In larger groups, people feel they do not have enough time to present their ideas, the leader becomes more autocratic and monopolizes the available time, and competing subteams may develop. Smaller groups often lack a definite leader, and they may not develop clear goals or may not have enough different perspectives to avoid groupthink.
The effectiveness of a team depends on the quality of the people in it and the coordination of its activities. However, how these variables impact on team effectiveness depends on the task. There are three types of tasks that need consideration:
Divisible versus Unitary Tasks. Divisible tasks can be done by different people—for example, checking the references in a reference list (one could divide the checking across as many clerks as there are pages of references to be checked). Unitary tasks cannot be divided—for example, understanding this paragraph.
Maximizing versus Optimizing Tasks. Maximizing tasks have a criterion with no limit—for example, find as many references as you can. Optimizing tasks have a criterion with an optimal level—for example, determine how much space you need for this project (too much space is wasteful and too little results in an ineffective project).
Disjunctive versus Conjunctive Tasks. In disjunctive tasks, if one member has the correct solution, the others will necessarily agree—for example, the root of a quadratic equation. In conjunctive tasks, every member must agree—for example, in a jury or in a committee in which everyone has veto power.
In disjunctive tasks, if the probability that one member has the correct solution is P, and the probability that no one can solve the problem is Q, where Q = 1 − P, the theoretical probability that the group will solve the problem is 1 − Qn, where n is the number of members. Clearly, the more members you have, the better the group's chances of being successful. In the case of conjunctive tasks, the opposite is true. A solution is more likely if the group is small than if the group is large.
Incidentally, empirically, groups do less well than the theory (Pg = 1 – Qn) predicts, primarily because there are losses in efficiency that occur when people discuss various solutions. Similarly, in divisible tasks, one would expect n people to produce n times the output of one person, but empirically this does not happen. The group often produces less. This lack of individual responsibility, called "social loafing," results when there is no clear identification of each person's output. Usually when one can identify the output of each individual, social loafing is small or nonexistent. Finally, unitary tasks are better done by individuals than by groups.
Osborn (1979) advocated the use of "brainstorming" to increase the creativity of small groups. According to his theory, when people generate ideas in the nonevaluative climate of a group and get stimulated from each other, they produce more ideas. However, the evaluation of this proposal has not supported this idea. For example, Dunnette and colleagues (1963) arranged for 48 research scientists and 48 advertising personnel either to do brainstorming or to work individually. In 23 of the 24 groups a greater number of different ideas were produced by the individuals than by the groups. Not only did individuals produce more ideas when working alone, but also the quality of these ideas was not inferior.
An important aspect of good group problem solving is the development of a wide range of alternatives. This can best be done when people think of a variety of courses of action independently of others. However, when it comes to evaluating these alternatives there is an advantage in having many different critics who examine the solution from as many points of view as possible. This critique also can be done individually, but groups can be more effective when the evaluation requires memory for previous events. Groups are better at remembering complex material than are individuals, who often forget some fact that others remember and can supply during the evaluation.
In forming effective R&D teams, leadership, individual problem-solving style, and work group relations affect innovative behavior directly and indirectly through their influence on perception of the climate for innovation (Scott, 1994). The results of this study also provide evidence that innovative behavior is related to the quality of the supervisor-subordinate relationship.
Currently a shift from hierarchically managed teams to self-managing teams is being seen in organizations. This change is partly due to the quality movement spreading throughout the United States. Concurrent engineering, which links the design, prototyping, and manufacturing process together, is a quality improvement trend that has also led to an increase in team work among technical professionals and managers (Levi and Slem, 1995). Self-managing teams reduce the need for supervisors and middle managers since responsibility is placed on the teams themselves.
A team is defined to be "a small number of people with complementary skills who are committed to a common purpose, set of performance goals, and approach for which they hold themselves mutually accountable" (Katzenbach and Smith, 1993, p. 112). Teams can either be temporary or permanent, but both types require interaction and the performance of common task. "Virtual teams" occur when group members interact without being located in the same place using mediums such as conference calls, electronic mail, and video conferences (Willis, 1991). Joint decision making among team members is used most often either through leader-controlled decisions or through a more democratic or consensus style of decision making.
Task, social relationships, and the individual are three areas Hackman (1987) uses to determine the success of a particular team. In looking at the success of a teamwork situation, the group should perform the task in a better and more efficient manner than an individual could do on his or her own. The next phase is social relationships or the building of skills and capabilities of the organization through the team members themselves. When this occurs, members will be empowered and encouraged to actively participate in the team, which leads to a more successful team environment. The individual is the third aspect for determining team success. An individual should be happy and his or her skills should be improved by working on a team with different skills and expertise (Katzenbach and Smith, 1993).
Research and development teams are unique from typical teams in that the tasks performed are nonroutine. There are many advantages for teams in R&D organizations. Usually R&D projects deal with complex tasks, requiring technical skills in multiple areas; consequently, there is an obvious need and benefits to using teamwork since tasks are challenging requiring diverse skills (Levi and Slem, 1995).
There are also some disadvantages to teamwork in the research and development enterprise. Problems can occur in the team setting simply because technical professionals are not selected for their social and communication skills, but for their scientific and technical skills. Many technical professionals seek this career so they will not have to work with others on a social level and research has shown that technical professionals may not have the traits needed in order to function well in a team environment (Lea and Brostrom, 1988). On another level, the reward system can also deter successful teamwork since many research and development professionals want to be rewarded for individual gains at work. Obviously teamwork is not an individual affair and therefore this may demotivate individuals; on the other hand, successfully completing complex research projects requires collaboration and teamwork: this may provide the needed motivation.
Based on this information, Levi and Slem (1995) decided to conduct a research project to determine the success of teams in research and development organizations. They developed a 30-item survey that measured the values and beliefs of the employees. The survey was distributed to three R&D facilities in the electronics industry in California. In addition to taking the survey, key professionals at each location were interviewed as a means of determining the validity of the surveyed information. Based on the analysis of this study, it has been found that research and development organizations can be good places to promote the use of teamwork. Lack of rewards in the team setting was viewed as a major problem from an individual perspective. This is because individuals do not often see the tie between success in a team environment and their individual success within the organization. It was also found that teams can be disrupted by members who cause conflict and disagreement, which in turn disturbs group cohesion and effectiveness. In addition, due to the variability in teams' characteristics and make-ups, there is no best approach to leadership and decision making. Sometimes organizations will promote a leadership style within a team that does not fit the nature of the project. This could lead to inefficiency and ineffective teams.
Self-managing teams are discussed at the beginning of this section and research has found that they are very difficult to create in an R&D setting. The three companies surveyed are all emphasizing the use of self-managing teams, though research has not proven this method to be superior over other teamwork types. Finally, Levi and Slem (1995) found that corporate culture affects the overall success of teamwork. Intuitively this makes sense. If an organization has a corporate culture that fosters employee involvement and participation, while emphasizing teams and their benefits, employees will be much more willing and able to be successful when put into a teamwork setting.