Although group discussion generally improves the quality of a group’s decisions, this will only be true if the group discusses the information that is most useful to the decision that needs to be made. One difficulty is that groups tend to discuss some types of information more than others. In addition to the pressures to focus on information that comes from leaders and that is consistent with group norms, discussion is influenced by the way the relevant information is originally shared among the group members. The problem is that group members tend to discuss information that they all have access to while ignoring equally important information that is available to only a few of the members, a tendency known as the shared information bias (Faulmüller, Kerschreiter, Mojzisch, & Schulz-Hardt, 2010; Reimer, Reimer, & Czienskowski (2010).
Research Focus: Poor Information Sharing in Groups
In one demonstration of the shared information bias, Stasser and Titus (1985) used an experimental design based on the hidden profile task, as shown in Table 10.1. Students read descriptions of two candidates for a hypothetical student body presidential election and then met in groups to discuss and pick the best candidate. The information about the candidates was arranged so that one of the candidates (Candidate A) had more positive qualities overall in comparison with the other (Candidate B). Reflecting this superiority, in groups in which all the members were given all the information about both candidates, the members chose Candidate A 83% of the time after their discussion.
Group member |
Information favoring Candidate A |
Information favoring Candidate B |
X |
a1, a2 |
b1, b2, b3 |
Y |
a1, a3 |
b1, b2, b3 |
Z |
a1, a4 |
b1, b2, b3 |
This is an example of the type of “hidden profile” that was used by Stasser and Titus (1985) to study information sharing in group discussion. Stasser, G., & Titus, W. (1985). Pooling of unshared information in group decision making: Biased information sampling during discussion. Journal of Personality and Social Psychology, 48(6), 1467–1478. (The researchers’ profiles were actually somewhat more complicated.) The three pieces of favorable information about Candidate B (b1, b2, and b3) were seen by all of the group members, but the favorable information about Candidate A (a1, a2, a3, and a4) was not given to everyone. Because the group members did not share the information about Candidate A, Candidate B was erroneously seen as a better choice. |
However, in some cases, the experimenters made the task more difficult by creating a “hidden profile,” in which each member of the group received only part of the information. In these cases, although all the information was potentially available to the group, it was necessary that it be properly shared to make the correct choice. Specifically, in this case, in which the information favoring Candidate B was shared, but the information favoring Candidate A was not, only 18% of the groups chose A, whereas the others chose the inferior candidate. This occurred because although the group members had access to all the positive information collectively, the information that was not originally shared among all the group members was never discussed. Furthermore, this bias occurred even in participants who were given explicit instructions to be sure to avoid expressing their initial preferences and to review all the available facts (Stasser, Taylor, & Hanna, 1989). Although the tendency to share information poorly seems to occur quite frequently, at least in experimentally created groups, it does not occur equally under all conditions. For one, groups have been found to better share information when the group members believe that there is a correct answer that can be found if there is sufficient discussion (Stasser & Stewart, 1992), and if they are forced to continue their discussion even after they believe that they have discussed all the relevant information (Larson, Foster-Fishman, & Keys, 1994). These findings suggest that an important job of the group leader is to continue group discussion until he or she is convinced that all the relevant information has been addressed. The structure of the group will also influence information sharing (Stasser & Taylor, 1991). Groups in which the members are more physically separated and thus have difficulty communicating with each other may find that they need to reorganize themselves to improve communication. And the status of the group members can also be important. Group members with lower status may have less confidence and thus be unlikely to express their opinions. Wittenbaum (1998) found that group members with higher status were more likely to share new information. However, those with higher status may sometimes dominate the discussion, even if the information that they have is not more valid or important (Hinsz, 1990). Groups are also likely to share unique information when the group members do not initially know the alternatives that need to be determined or the preferences of the other group members (Mojzisch & Schulz-Hardt, 2010; Reimer, Reimer, & Hinsz, 2010). Findings showing that groups neither share nor discuss originally unshared information have very disconcerting implications for group decision making because they suggest that group discussion is likely to lead to very poor judgments. Not only is unshared information not brought to the table, but because the shared information is discussed repeatedly, it is likely to be seen as more valid and to have a greater influence on decisions as a result of its high cognitive accessibility. It is not uncommon that individuals within a working group come to the discussion with different types of information, and this unshared information needs to be presented. For instance, in a meeting of a design team for a new building, the architects, the engineers, and the customer representatives will have different and potentially incompatible information. Thus leaders of working groups must be aware of this problem and work hard to foster open climates that encourages information sharing and discussion. Given its obvious pitfalls, an interesting question to ask is why the shared information bias seems to be so pervasive. Recalling the confirmation bias that we discussed in the chapter on social cognition, perhaps it reflects this tendency played out at the group level, where group members collaborate to provide confirmatory evidence for each other’s positions. Leading on from this, it could also reflect the tendency for people to wish to use groups to reinforce their own views. Perhaps sometimes groups become places where people seek to mutually validate each other’s shared perspectives, to the detriment of them searching out the alternatives. If these ideas are correct, given that we often choose to associate with similar others, then it may be important to seek out the views of group members that are likely to be most different from our own, in seeking to weaken the damaging effects of the shared information bias (Morrow & Deidan, 1992).
- 3613 reads