Why Groups Make Great and Terrible Decisions

Lecture Overview: Group Decision-Making Processes

  • The lecture, presented by Prof Tim Hopthrow, explores the dynamics of how groups make decisions, categorized into three primary inquiries:   - Does interaction with groups intensify our decisions?   - Does unique knowledge contribute to group decisions?   - Do groups make better decisions than individuals?

Group Polarization and the Risky Shift

  • Part 1 Question: Does interaction with groups intensify our decisions?

  • Group Polarization Definition: A tendency for group discussions to produce more extreme group decisions than the individual members' pre-discussion opinions or the direction originally favored by the mean.

  • Risky Shift (Stoner, 1961):   - In this study, individuals initially marked their answers alone.   - After discussing opinions with a group, they marked their answers again.   - Findings: Groups consistently made riskier decisions than individuals. The group consensus shifted toward a "riskier" position than the average decision made by individuals prior to the discussion.

  • Re-conceptualization (Moscovici & Zavalloni, 1969):   - Proposed that the "Risky Shift" is not specifically about risk, but is a general phenomenon related to the group's initial leanings.   - Group Polarization: Group discussion strengthens the average inclination of the members.

  • Moscovici & Zavalloni (1969) French Student Study:   - Examined attitudes toward Americans (initially negative) and toward General de Gaulle (initially positive).   - Post-discussion results:     - Attitudes toward Americans became even more negative.     - Attitudes toward de Gaulle became even more positive.     - Data trends showed a shift from initial scores (approx. 1.0-1.0 for Americans and 1.01.0 for de Gaulle) to more extreme scores (approx. 1.5-1.5 and 1.51.5 respectively).

  • Further Evidence: Brauer, Judd, & Jacquelin (2001) found that mutual dislike for peers increased when students shared those views with one another.

Mechanisms of Group Polarization

  • 1. Normative Influence (Social Comparison):   - Driven by the desire for public approval.   - Individuals adjust their views to more extreme positions to fit in or be seen favorably by the group.

  • 2. Informational Influence (Persuasive Argument):   - Verbalizing ideas that support the dominant opinion reinforces belief in that opinion.   - Hearing new persuasive arguments in favor of the dominant view further polarizes beliefs.

  • 3. Self-Categorization Theory:   - Identifying as a group member leads individuals to take on the identity of intergroup behaviors.

The "Groupdrink" Study

  • Are groups always risky? Research by Abrams, Hopthrow, Hulbert, and Frings (2006) examined risk orientation under the influence of alcohol.

  • Study Design: Evaluated risk orientation in groups vs. alone, and after consuming alcohol vs. a placebo.

  • Findings:   - Main effect of alcohol/placebo: Alcohol lead to higher risk attraction than the placebo (Alcohol > Placebo).   - Main effect of group/individual: None found.   - Significant interaction effect: For individuals, alcohol increased risk significantly (from approx. 2.22.2 to 4.04.0). However, for groups, the increase in risk from placebo to alcohol was much smaller (approx. 3.23.2 to 3.63.6).   - Conclusion: This supports the idea of "group monitoring"; groups may exercise more caution or monitoring when members have been drinking compared to individuals.

The Hidden Profile Problem

  • Part 2 Question: Does unique knowledge contribute to group decisions?

  • Basic Concept: Groups possess collective shared knowledge and unique individual knowledge.

  • Theoretical Perspective (Stasser & Titus, 1985, 1987): Initially suggested that unique and novel arguments should be the most persuasive in decision-making groups.

  • Empirical Reality (The Hidden Profile Problem): Groups generally focus on information shared by all members rather than information uniquely held by individuals (Stasser & Titus, 1985, 1987, 2003).

  • Key Characteristics:   - Groups over-focus on common insights and fail to discuss unique information (Strasser, 1992).   - Even when shared, unique information is often not heard, understood, or integrated (Dennis, 1996; Kerr & Tindale, 2004; Larson et al., 1994).   - This failure leads to lower quality decisions.

Factors Causing the Hidden Profile

  • Time Pressure: High urgency for closure makes participants less receptive to diverging perspectives, though they become more task-focused (Kruglanski & Webster, 1991; Karau & Kelly, 1992).

  • Group Size: Larger groups increase the likelihood of overlooking unique ideas (Steiner, 1972).   - In a 33-person group: Each individual has a 33.3%33.3\,\% chance to contribute unique info.   - In a 44-person group: Each individual has a 25%25\,\% chance to contribute unique info.

  • Diversity: Greater diversity can lead to intergroup stereotypes and conflicts that compromise member resources (Caruso & Woolley, 2008).

  • Composition Shifts: Frequent changes in group membership make it hard to track who has specific expertise (Lewis et al., 2007).

  • Technology/Dispersion: Geographically dispersed groups relying on electronic communication struggle to determine "who knows what" (Nemiro et al., 2008).

  • Social Costs & Status:   - High-status members (e.g., resident physicians) are more likely to discuss unique info than low-status members (e.g., interns/students) due to credibility concerns for the latter (Larson et al., 1996; Stewart & Stasser, 1995).

Strategies to Overcome Hidden Profiles

  • Groups must recognize members with special expertise (Bauman & Bonner, 2004).

  • The group must value individual contributions (Caruso & Woolley, 2008).

  • Motivation and coordination are required to utilize resources (Quigley et al., 2007).

  • Ensure equal status among members (Hollingshead, 1996).

  • Use common dimensions for comparison (e.g., ranking decisions rather than choosing outright) (Reimer et al., 2007).

Brainstorming in Groups

  • Part 3 Question: Do groups make better decisions than individuals?

  • Definition: The generation of novel ideas where members generate as many ideas as possible, as quickly as possible (Osborn, 1957).

  • Effectiveness: Diehl and Stroebe (1987) found little evidence that group brainstorming is more creative than individual brainstorming.

  • Exceptions & Nuances:   - Eyewitness testimony is more accurate as a group account (Warnick & Saunders, 1980).   - Groups produce better quality ideas when they critique each other (McGlynn et al., 2004).

  • Factors hindering brainstorming (Paulus et al., 1993):   - Evaluation apprehension.   - Social loafing and free riding.   - Production matching (matching the low output of others).   - Production blocking (forgetting ideas while others speak).

  • Improving Brainstorming:   - Combine individual and group sessions (Brown & Paulus, 2002).   - Use written communication (Mullen et al., 1991) or electronic brainstorming (Gallupe et al., 1994).   - Use heterogeneous groups with diverse backgrounds (Stroebe & Diehl, 1994).

  • Illusion of Group Effect: People believe groups are better because a group generates more total ideas than any single individual, even though an equal number of individuals working alone generate more total ideas than the group (Diehl & Stroebe, 1991).

Group Memory and Transactive Memory

  • Group Memory: Groups generally remember more than individuals (Clark & Stephenson, 1989, 1995). Logic suggests they recognize true information when heard (Lorge & Solomon, 1995).

  • Complexity Factor: Steiner (1976) noted groups excel at simplistic artificial tasks (e.g., nonsensical words) more than complex realistic tasks (e.g., a story).

  • Transactive Memory (Wegner et al., 1991):   - A shared system for encoding, storing, and retrieving information that exists only within a group.   - Example: A psychology department or a couple where different individuals are responsible for different domains (e.g., car insurance vs. domestic chores).

  • Group Mind (McDougall, 1920): Theory that people adopt a qualitatively different state of mind when in a group.

  • Stability in Transactive Memory (Argote et al., 1995):   - Study involved 240240 people in groups of 33 learning origami.   - Performance dropped as members left; newcomers did not improve performance because they hadn't learned with the group.

Groupthink

  • Definition (Janis, 1972): A desire to reach a unanimous decision that overrides the motivation for a rational decision-making process.

  • Vulnerability Factors: Similar backgrounds, insulation from outside opinions, and lack of clear decision rules.

  • Eight Symptoms (Janis, 1971):   - 1. Illusion of invulnerability (excessive optimism).   - 2. Collective rationalization (discounting warnings).   - 3. Belief in inherent morality (ignoring ethical consequences).   - 4. Stereotyped views of out-groups (viewing "enemies" as ineffective).   - 5. Direct pressure on dissenters.   - 6. Self-censorship (doubts are not expressed).   - 7. Illusion of unanimity (assuming the majority view is unanimous).   - 8. Self-appointed ‘mindguards’ (protecting the leader/group from contradictory info).

Remedies and Critiques of Groupthink

  • Remedies for Groupthink (Janis, 1982):   - Assign a "devil's advocate"/critical evaluator role.   - Leaders should avoid stating preferences early.   - Discuss deliberations with outside trusted associates.   - Invite outside experts to challenge decisions.   - Have a "second chance" meeting to discuss lingering doubts.

  • Critiques:   - Research is often retrospective.   - Inconsistent findings: while high cohesion can lead to groupthink (Rovio et al., 1996), it doesn't always. Norms of critical thought can lead to correct decisions even in cohesive groups (Postmes, Spears & Cihangir, 2001).