<<

. 59
( 115 .)



>>

ual group members will deindividuate. Differences have been found in the behavior of
larger and smaller crowds that gather when a troubled person is threatening to leap from
a building or bridge (Mann, 1981). Out of 21 such cases examined, in 10, the crowds
baited the victim to jump, whereas in the remaining 11, the victim was not taunted and
was often rescued. What was the difference between these two sorts of cases?
The baiting crowds tended to be larger”over 300 people. The baiting episodes
were more likely to take place after dark, and the victim was usually situated higher
up, typically above the 12th ¬‚oor. Additionally, the longer the episode continued, the
Social Psychology
296

more likely was the taunting. All these factors”the large size of the crowd, the distance
between that crowd and the victim, the anonymity lent by darkness”contributed to
the deindividuation of the members of the crowd. And the longer these deindividuated
people waited, the more irritable they became.
Another study found that when a crowd is bent on violence, the larger the crowd,
the more vicious the behavior (Mullen, 1986). Larger crowds and smaller numbers of
victims can lead to atrocities such as hangings, torture, and rape.



Group Performance

Individual Decisions and Group Decisions
First of all, letʼs consider whether group decisions are in fact better than individual
decisions. Is it better to have a team of medical personnel decide whether our CAT
scan indicates we need surgery, or is that decision better left to a single surgeon? Did
the launch director at NASA bene¬t from the workings of the group, or would he have
been wiser to think through the situation on his own?

Does a Group Do Better Than the Average Person?
In general, research shows that groups do outperform individuals”at least the average
individual”on many jobs and tasks (Stasser, Kerr, & Davis, 1989). Three reasons
have been proposed for the observed superiority of groups over the average person.
First of all, groups do a better job than the average person because they recognize
truth”accept the right answer”more quickly. Second, groups are better able to reject
error”reject incorrect or implausible answers (Laughlin, 1980; Laughlin, VanderStoep,
& Hollingshead, 1991; Lorge & Solomon, 1955). Third, groups have a better, more
ef¬cient memory system than do individuals. This permits them to process informa-
tion more effectively.
However, groups do not appear to live up to their potential. That is, their performance
seems to be less than the sum of their parts (i.e., the individual members [Kerr & Tindale,
2005]). So letʼs keep that in mind as we ¬rst see what advantages groups have over
transactive memory individuals. Groups may possess what has been called transactive memory systems,
systems Systems within a shared system for placing events into memory (encoding), storing those memories,
groups that are sets of and retrieving that information. Wegner (1996) used the example of a directory-sharing
individual memories that allow
computer network to explain the three legs of a transactive memory system:
group members to learn about
each other™s expertise and
1. Directory updating, in which people ¬nd out what other group members know
to assign memory tasks on
that basis.
2. Information allocation, the place where new information is given to the person
who knows how to store it
3. Retrieval coordination, which refers to how information is recovered when
needed to solve a particular problem
Group members learn about each otherʼs expertise and assign memory tasks on that
basis. This not only leaves others to concentrate on the memory tasks they do best,
it also provides the group with memory aids. Someone in the group may be good in
math, for example, so that person is assigned the task of remembering math-related
information. When the group wants to recall that information, they go to this expert
Chapter 8 Group Processes 297

and use him or her as an external memory aid. Memory thus becomes a transaction,
a social event in the group. For some or all of these reasons, groups seem to outper-
form the average person on many decision-related tasks (Laughlin, Zander, Knievel,
& Tan, 2003).
Hollingshead (1998) showed the effectiveness of transactive memory. She studied
intimate couples as compared to strangers who worked on problems, some face to face
and others via a computer-conferencing network. Intimate couples who were able to sit
face-to-face and process their partnerʼs verbal and nonverbal cues were able to solve
problems better than couples comprised of strangers, because the intimate couples
were able to retrieve more information. Intimate couples who worked via a computer-
conferencing system did not do as well, again suggesting that the nonverbal cues were
important in pooling information. In fact, recent research shows that in small groups
in which the individual members do not submerge their personal identities but rather
express them, the individualsʼ identi¬cation with that group is enhanced (Postmes,
Spears, Lee, & Novak, 2005).

Does a Group Do Better Than Its Best Member?
We noted that research shows that groups outperform the average person. But does the
group perform better than the best member, the smartest person, the “best and bright-
est” member of the group?
To test the hypothesis that groups can ¬nd correct responses better than individu-
als, college students were asked to try to discover an arbitrary rule for separating a
deck of cards into those that did and did not ¬t the rule (Laughlin, VanderStoep, &
Hollingshead, 1991). If the rule was “hearts,” for example, then all cards of the hearts
suit would ¬t the rule, and all others would not. Subjects had to guess the rule, and
then test it by playing a card. The feedback from the experimenter gave them infor-
mation on which to base their next guess. The researchers also varied the amount of
information that subjects had to process. They presented some subjects with only two
arrays of cards, others with three, and others with four: The more arrays, the more
dif¬cult the task.
The performance of four-person groups was then compared to the performance of
each of the four group members, who had to do a similar task individually. The best indi-
vidual was able to generate more correct guesses than the group or any other individual
member. The groupʼs performance was equal to its second-best member. The third- and
fourth-best members were inferior to the group. As the task became more dif¬cult”the
arrays increased to four, which made much more information available”the perfor-
mance of both the best individual and the group fell. The researchers also compared the
abilities of groups and their individual members in rejecting implausible hypotheses.
The fewer implausible ideas subjects or groups raised, the better they did with respect
to rejecting false leads. Groups and the best individual were better at rejecting false
leads than were the second-, third-, and fourth-best individuals.
This research suggests that groups in general perform as well as their best or second-
best individual member working independently. You might ask, Why not just let the
best member do the task? But keep in mind that it is often not possible to identify the
groupʼs best member prior to completing the task. This ¬nding tells us that groups tend
to perform competently, particularly when the information load is not overwhelming.
In addition, it may very well be that the kind of problem that the group has to deal
with may in¬‚uence whether or not a very good individual is or is not better than the
group solution.
Social Psychology
298


The Harder the Problem, the Better the Group
Recent work suggests that we may have underrated the ability of groups to reach solu-
tions, especially more dif¬cult problems. Crott, Giesel, and Hoffman (1998) argued that
their research on group problem solving suggests that dif¬cult tasks provoke creativity
in groups. When faced with a problem that required the group to come up with a number
of hypotheses to discover the correct answers, groups more than individuals were able
to generate a number of novel explanations. Groups were also shown to be less likely
to be prone to the con¬rmation bias than were individuals (Crott et al., 1998).
Similarly, Laughlin, Bonner, and Altermatt (1998) showed that groups were as good
as the best individual in solving dif¬cult inductive (proceeding from speci¬c facts to
general conclusions) problems and better than all the remaining group members. Groups
are especially effective in dealing with information-rich problems because they have
more resources (Tindale, Smith, Thomas, Filkins, & Sheffey, 1996).
The ¬nding that the best member of a group may outperform the group is also modi-
¬ed by the size of that group and by the type of problem. Laughlin and his colleagues
studied groups that varied in size from two to ¬ve people (Laughlin, Hatch, Silver, &
Boh, 2006). The groups had to deal with a complex intellectual problem that required
different strategies. The researchers ¬rst determined the best, second-best, third-best,
and fourth- and ¬fth-best member of each group. Laughlin et al. then compared the solu-
tions to these complex problems submitted by individual members and those submit-
ted by three-, four-, and ¬ve-person groups. These researchers found, contrary to some
previous ¬ndings, that the groups took signi¬cantly less time to solve problems and the
quality of the solutions were better than those of the best member of the group. That
is, each of the three-, four-, and ¬ve-person groups solved the problems more quickly
and produced more complex solutions to the problems than the best individual member.
And, there were no signi¬cant differences between three-, four-, and ¬ve-person groups.
This is interesting because we might have expected some “motivation loss” due to free
riders (see our earlier discussion) as the group got larger.
What about the two-person groups? The two-person groups performed less well
than the other groups. Laughlin et al. (2006) concluded that groups of three that are
“necessary and suf¬cient” perform better than the very best individual on dif¬cult intel-
lective problems.
We have seen how well groups perform with respect to the abilities of their
members. Letʼs take a closer look at the workings, the dynamics, of how those deci-
sions are made.
How do groups gather and use the information possessed by individual members?
How do they reach decisions?

The Groupʼs Use of Information: Hidden Pro¬les
One advantage groups have over individual decision makers is that a variety of indi-
viduals can usually bring to the discussion a great deal more information than can one
person. This is usually seen as the great advantage of groups. But does the group make
adequate use of that information? Research shows that group members tend to discuss
information that they share and avoid discussing information that only one person has.
This research on the insuf¬cient sharing of information that one member of the group
may have is known as the hidden pro¬le paradigm. The hidden pro¬le paradigm refers
to a situation in which the groupʼs task is to pick the best alternative, say the best job
applicant, but the relevant information to make this choice is distributed among the
group members such that no one member has enough information to make the right
choice alone (Greitmeyer & Schulz-Hardt, 2003).
Chapter 8 Group Processes 299

In one experiment, each member of a committee received common information
about three candidates for student government (Stasser & Titus, 1987). Each also
received information about each candidate that none of the others received (unshared
information). The committee members met in four-person groups to rank the candi-
dates. The sheer number of facts available to the members varied from one group to
the next. When the number of facts was high, the raters ignored information that was
unshared. That is, they rated the candidates based solely on the information that they
held in common. The information they chose to share tended to support the group
decision; they did not share information that would have con¬‚icted with the decision.
Because the results of this study indicate that group members try to avoid con¬‚ict
by selectively withholding information, the researchers concluded that face-to-face,
unstructured discussion is not a good way to inform group members of unshared infor-
mation (Stasser, 1991).
There appear to be at least two reasons for the failure of face-to-face groups to report
and use unshared information. The ¬rst has to do with the way people think. Whatever
is most salient (the shared information) tends to overwhelm that which recedes into
the background (the unshared information). In other words, group members hear the
shared information and simply neglect to bring up or take into account the unshared
information. The second reason is that individuals may be motivated to ignore or forget
information (unshared) that they think may cause con¬‚ict. Individuals also avoid dis-
cussing or disclosing information that goes counter to the groupʼs preferred decision
(Greitemeyer & Schulz-Hardt, 2003).
The nature of a groupʼs task may also affect how the group searches for informa-
tion and uses shared and unshared facts. To investigate this possibility, experimenters
hypothesized that groups would be more likely to share all information if they knew
that the problem had a de¬nitively correct answer than if the task called only for a
judgment (Stasser & Stewart, 1992). Subjects in this study were given information
about a crime. In some groups, all the information was given to all the members. In
other groups, some information was given only to individual members. In other words,
in the latter groups, some members had unshared information. In addition, half the
groups were told that there was enough evidence to solve the crime, whereas others
were informed that because the evidence was less than full, the group would have to
make a judgment call.
The results showed that groups given the task with the correct answer were much
more likely to search for the unshared information and get the right answer than groups
given a judgment problem. What differed was the expectation that there was or was not
a correct answer (Stasser & Stewart, 1992). When the group members think or know
that the task has a de¬nite answer, they are more forthright in bringing up anything
(unshared) that could help the group. The group strategy changes because people want
to search for any information that helps them to be successful. Greitmeyer and Schulz-
Hardt (2003) have shown that if a hidden pro¬le has incorrect information, you are
unlikely to detect that error. If you do not share your hidden pro¬le with others, then it
is improbable that the error would be recti¬ed.
The research of James R. Larson, Jr., showed that access to unshared information
is crucial to good group decision making. For example, Larson, Christensen, Franz, and
Abbot (1998) examined the decision making of medical teams. Three-person physi-
cian teams had to diagnose cases and were given shared information (to all three MDs),
whereas the rest of the diagnostic data were divided among the three. Compared with
Social Psychology
300

unshared information, the physicians discussed shared information earlier in the dis-
cussion. However, the unshared information, when discussed, proved to lead to more
accurate (correct diagnosis) outcomes.
In other research, Larsonʼs team reached similar conclusions. Winquist and Larson
(1998) gave three-person groups the task of nominating professors for teaching awards.
Discussion focused more on shared information, but the quality of the decision was
determined by the amount of unshared information that was pooled in the discus-
sion (Henningsen, Dryden, & Miller, 2003). One way to increase the likelihood that
unshared hidden pro¬les will be brought to the discussion is to suggest to the group
members that they think in a counterfactual way. That is, if you have some information
that nobody else has, you might say “What if this is inaccurate, what would it mean?”
If that is done, it seems to be the case that more unshared information sees the light of
day (Galinsky & Kray, 2004).

The Effect of Leadership Style on Group Decision Making
How can we make sure groups gain access to unshared information? What is the best
way of making sure that group members who have information that others do not are
motivated to pool that information?
We know that leadership style is important in determining how groups function
(Fiedler, 1967). In one study, researchers identi¬ed two common styles of leadership.
participative leader The ¬rst, the participative leader, shares power with the other members of the group
A leadership style and includes them in the decision making. Another leadership style, the directive leader,
characterized by a leader gives less value to participation, emphasizes the need for agreement, and tends to prefer
who shares power with the
his or her own solution.
other members of the group
and includes them in the
Directive and Participative Leaders
decision making.
Research using these leadership styles indicated that participative leaders provoked their
directive leader
groups to discuss more information, both shared and unshared, than did groups with a
A leadership style involving a
leader who gives less value directive leader (Larson, Foster-Fishman, & Franz, 1998). However, directive leaders
to participation, emphasizes were more likely to repeat information that had been pooled, especially unshared infor-
the need for agreement, and mation. In other words, directive leaders made unshared information more prominent.
tends to prefer his or her own
It seems, then, that participative leaders worked to get the group to bring out more
solution.
information but that directive leaders were more active in managing the information
once it was put on the table. What about the quality of the decisions? Interestingly,
groups under participative leadership made many more incorrect decisions. This was
counter to the researchersʼ expectations (Larson et al., 1998). If directive leaders have
information that favors the best alternative, they use it and bring the group to a good-
quality decision. They do this much better than participative leaders. The downside to
directive leaders is that they may not be able to get the group members to bring out all
the necessary information for good decision making.

Gender and Leadership
Eagly and her colleagues have investigated the possible differences in leadership styles

<<

. 59
( 115 .)



>>