4956 teamwork

groups, focus on real world groups has
typified more recent research in this area.
Some studies suggest facilitative effects of
heterogeneity on team performance. This pat
tern is most likely when the characteristics in
question are skills or educational specialization.
Strategic management initiatives appear more
likely to be made by groups that were heteroge
neous with respect to educational specialization.
More recently, studies in health care suggest that
the greater the number of professional groups
represented in teams, the higher the levels of
innovation in patient care. It might be that skill
heterogeneity means that each group member is
more likely to have non redundant and, pre
sumably, relevant expertise to contribute to the
team activities. Groups that include both diverse
and overlapping knowledge domains and skills
are particularly creative.
Some debate has surrounded the question of
whether it is advantageous to have groups that
are homogeneous or heterogeneous with respect
to cognitive ability. Results of two recent meta
analyses suggest that the relation between abil
ity heterogeneity and performance may be
somewhat complex. Based on these analyses, it
appears that, in general, ability heterogeneity
and performance are unrelated. Thus, there
would seem little justification to select team
members with a view to dispersing their cogni
tive ability levels.
Teams which are diverse in task related attri
butes are often diverse in relation to attributes
inherent in the individual. These relation
oriented characteristics can trigger stereotypes
and prejudice which, via intergroup conflict
(Hogg & Abrams 1988), can affect group pro
cesses and outcomes. For example, turnover
rates are higher in groups that are heterogeneous
with respect to age. Two studies that have exam
ined ethnicity diversity in groups have sug
gested that the effects of diversity may change
over time. Milliken and Martins (1996) sug
gested that ethnic diversity in groups can have
negative effects on individual and group out
comes, primarily early in a groups life. Simi
larly, in one of the very few longitudinal studies
in this area, Watson et al. (1993) reported that
groups that were heterogeneous with respect
to culture initially performed, on a series of
business case exercises, more poorly than cultu
rally homogeneous groups. As group members
gained experience with each other over time,
however, performance difference between cul
turally homogeneous and heterogeneous groups
largely disappeared.
Organizational supports. Various organiza
tional contextual factors have been proposed as
important in predicting team effectiveness.
Reward systems, such as public recognition, pre
ferred work assignments, and money, have long
been known to provide motivation and affect
performance, particularly when the rewards are
contingent upon task achievement. Gladstein
found that pay and recognition had an effect,
especially upon the leaders behavior and the
way the group structured itself. Hackman
(1990) identified two contingencies: whether
the rewards are administered to the group as a
whole or to individuals, and whether the rewards
provide incentives for collaboration or delega
tion of tasks to individuals (the former, in both
cases, are associated with positive relationships
between rewards and group effectiveness). Feed
back is important for setting realistic goals and
fostering high group commitment. In addi
tion, high job satisfaction requires accurate feed
back from both the task and other group
members. However, group feedback can be dif
ficult to provide to teams with either long cycles
of work or one off projects. Limited empirical
evidence suggests training is correlated with
both self reported team effectiveness and man
agers judgments of effectiveness.

Team Processes

The second major element of the inputpro
cessoutput model is team processes. Among
these, the most consistently important factor in
determining team effectiveness is the existence
of team goals or objectives (Guzzo & Shea 1992).
Objectives. The clarity or specificity of goals
has also been shown to predict team perfor
mance outcomes. In order to combine efforts
effectively, team members have to understand
collectively what it is they are trying to achieve.
Much research also indicates that involvement in
goal setting fosters commitment to those goals
and consequently better team performance.
Participation. The second factor of central the
oretical and empirical concern in the study of
team performance is the notion of participation.