Two well-known contributors in the team decision making field recently published a book on best practices (Sunstein & Hastie, 2015). Sunstein and Hastie argued that current techniques in team decision making have neglected research and, as a result, have a tendency to end unfruitfully. This neglect is problematic, as research shows that teams can potentially outperform individuals and there are tasks that only cross-functional teams with complementary skills can perform (Hinz, Tindale, & Vollrath, 1997; Kerr & Tindale, 2004; van Ginkel, Tindale, & van Knippenberg, 2009; Kameda, Tsukasaki, Hastie, & Berg, 2010).
The proliferating use of team-based structures has left many in the current workforce with important decisions to make, but potentially faulty processes to make them. The lack of general best practices is an issue for organizations, particularly when considering the impact of many top management teams that are responsible for leading entire organizations. To build a better understanding of best practices, the current paper focuses on three prominent factors of team decision making. These factors include information sharing, team diversity, and politics. With increased knowledge and fervor about the importance of this line of research, the neglected connection between research and application may advance.
To begin, there are multiple theorized models for team decisions and the aggregation of individual choices into a single team choice. For instance, there is the Teams Lens Model (Brehmer & Hagafors, 1986), the Multilevel Theory of Team Decision Making (Hollenbeck et al., 1995; Phillips, Douthitt, & Hyland, 2001), the Vroom-Yetton Contingency Model (1973), and the Judge-Advisor Systems (Sniezek & Buckley, 1995). These models are diverse, but team decisions are not simple procedures. There are multiple approaches include majority rules, rank voting, plurality, to requirements for unanimity.
The Multilevel Theory of Team Decision Making is perhaps the most relevant within the current characteristics of the American workforce. This approach acknowledges the status differences that occur within teams and the unequal distribution of knowledge and expertise between members. Influencing factors include the social environment, role, individual, tasks, behavioral settings, physical/technical environment, dyadic relationships, team information, and hierarchies within the team. The work of Hollenbeck and colleagues was important for team decision making research because it emphasized that team processes are similar to individual decisions and there are numerous factors affecting the decision making process. The Double Edged Sword of Information Sharing
One of the most highly researched areas in team decision making relates to information sharing and the hidden profile paradigm. Stasser and Titus created the concept of the hidden profile in 1985. In 1988, Stasser defined the hidden profile as an unmeasurable process involving shared and unshared information that contributes to biased decisions. Deciding whether to share information, particularly unshared novel information, dictates the hidden profiles that exist within a team.
As a number of meta-analyses illustrate (e.g., Lu, Yuan, & McLeod, 2012; Mesmer-Magnus & DeChurch, 2009; Reimer, Reimer, & Czienskowski, 2010), hidden profiles have been associated with poorer team decision outcomes. In practice, teams that share information often have better opportunities for making effective decisions. Teams that share information are thought to operate under a manifest profile (Lavery, Franz, Winquist, & Larson, 1999). Several researchers have demonstrated that identifying the hidden profile can be difficult for teams and many often fail (Wittenbaum, Hollinshead, & Botero, 2004; Strasser & Titus, 1985; Brodbeck, Kerschreiter, Mojzisch & Schulz-Hardt, 2007).
The research on hidden profiles and information sharing is interesting because it presents information sharing and the discussion of initial preferences as a double edge sword. Greitemeyer and Schulz-Hardt (2003) believed that dissent of information (i.e., discussion of initial preferences) and hidden profiles can deter the decision making process. They contended that individuals would be more likely to evaluate new and unfamiliar information more critically and perceive it as false when compared to information that was more familiar. They referred to this process as cognitive economy.
Their findings indicated that when teams are given no information they actually perform better than teams given initial information with a developed initial preference. Wittenbaum, Hollingshead, and Botero (2004) completed a review of hidden profiles and agreed with Greitemeyer and Schulz-Hardt. This issue relates back to the shared information effect (also referred to as the common information effect or knowledge effect) and a potential confirmation bias. The shared information effect refers to the principle that what is familiar with the team is usually discussed at greater lengths than the unshared information (Wittembaum & Stasser, 1996; Larson, Foster-Fishman, & Keys, 1994). In fact, research has shown that unshared details are ignored at the onset and better decisions are often not accomplished because they were not as apparent (Gigone & Hastie, 1993, 1997).
In addition, there is research to show that premature preference discussions influence the identification of the hidden profiles (Gigone & Hastie, 1993, 1997). Gigone and Hastie (1997) showed that teams that discussed final decisions focused more on the initial information, rather than novel facts. Knowing others’ initial preferences can potentially damage team decisions as initial dissent leads to favored decisions that negatively affect the quality of decisions (Mojzisch & Schulz-Hardt, 2010). This is a prime example of the evaluation bias or the confirmation bias, which states that individuals view their information as more relevant than contradictions to their decisions (Nickerson, 1998).
On the other hand, studies have shown that dissent of information can help identify the hidden profile and improve decisions (Schulz-Hardt, Brodbeck, Mojzisch, Kerschreiter, & Frey, 2006). For instance, Schulz-Hardt and colleagues showed that when team members disclosed partial or even full information on their preferences, the quality of the decision and discussion were positively impacted. Interestingly, the positive impact held true even when the initial preferences were incorrect.
Teams that held incorrect beliefs at the beginning of the discussion, but disclosed information were likely to identify their errors. The diverse preferences teams engaged in more quality discussions, such that the teams had fewer overall biases than teams lacking diversity. The authors believed that the discussion of preferences and information on the onset served as a helpful tool for the team.
In a review of team performance and information sharing, Mesmer-Magnus and DeChurch (2009) found that teams share information when the task demonstrability, discussion structure, information independence, cooperation in discussion, and member similarity are high. Furthermore, information sharing provides team members an opportunity to learn, to increase shared understanding of team tasks, and to streamline the management of behaviors (Cannon-Bowers, Tannenbaum, Salas, & Volpe, 1995). There is further support for the importance of cognition and affect during the information sharing process. Mojzisch and Schulz-Hardt (2010) showed that the process of encoding and integrating information was one underlying mechanism.
They used a team-based experiment focusing on initial discussion. Results showed that dissent deterred the teams from identifying the hidden profile, as members were unlikely to listen to information shared and less likely to encode information. Furthermore, they contended that the teams were more likely to identify the hidden profile when told not to provide initial preferences. Emich (2014) demonstrated that both positive and negative affect influenced information sharing. Teams with higher member level positive affect were more likely to share information when compared to negative affect teams. On the contrary, van Knippenberg, Kooij-de Bode, and van Ginkel (2010) found interactive effects that showed that teams high in positive mood, but low in trait negative affect, made poorer decisions.
Unfortunately, there are multiple hindrances to sharing information and initial preferences. van Ginkel et al. (2009) hypothesized that team member’s often refrain from distribution of information because of a failed understanding of the team’s tasks. To combat this failed understanding, they believed that teams should engage in team reflexivity and information elaboration to improve decision making. Reflexivity is the process of discussing tasks, objectives, and strategies to reach high performance. Information elaboration, much as it sounds, is the team member process of distributing, assimilating, and conferring about all information. van Ginkel and colleagues demonstrated that reflexivity encouraged common task understanding and talking over the task and reflecting helped individuals to make better decisions.
They also showed that when teams make quick decisions or when elaboration of information is not expected to influence the quality of the decision, reflexivity could in fact harm performance. Nonetheless, in the majority of teams, reflection leads to increased quality of decisions in longer tasks. Their findings provided additional support for Wittenbaum and colleagues’ (2004) argument that it is not simply what information is shared, but also how the information is communicated that is important as it could potentially backfire.
There is also the issue of team-serving biases, such as the “there is no I in team.” A team-serving bias explains why many individuals who make judgments towards their own teams will make positively based decisions (Sherman & Kim, 2005). Individuals tend to overly identify with their team and often defend their team as a means of protecting themselves, which stems from Social Identity theory (Tajfel & Turner, 1979, 1986). Sherman and Kim (2005) also attributed the team-serving bias to Self-Affirmation theory (Steele, 1988). Self-Affirmation theory explains how individuals react when their self-image is threatened.
Specifically, people are motivated to maintain a positive image and reduce possible threats. Individuals who are more affirmed in themselves can feel less threatened because they are more secure and less concerned about contradictions (Correll, Spencer, & Zanna, 2004; Cohen, Aronson, & Steele, 2000). It is very likely that the outcomes that affect team performance are likely to influence an individual and their perceptions of their self-image. The process would seem to be cyclical. Team decisions affect the individual, which in turn leads an individual to make judgments about the team. Team Diversity and Conflict
The second factor related to decision making is the diversity within the team. There is an abundance of research concerning team diversity, the cognitive judgments individuals make, and subsequent performance. It is important to note, diversity can be defined in a number of ways. Diversity can be defined based on demographic factors (often referred to as social categorization) or simply based on diversity of ideas (often referred to as information/decision making diversity). The bulk of team diversity research has focused on social categorization and functional backgrounds (van Knippenberg, De Dreu, & Homan, 2004). Most research suggests that diversity in teams can have both a positive and negative effect on performance (Guzzo & Dickson, 1996; Milliken & Martins, 1996; Williams & O’Reilly, 1998; van Knippenberg et al., 2004). Research has argued that differences in social categories tend to harm performance, but differences in information and perspectives tend to improve performance (Wit, Greer, & Jehn, 2012)
Diversity in teams is important because individuals tend to develop cognitive schemas about those around them. They place others with similar social categories into the “in-team” category, but treat dissimilar individuals as “out-team” members creating hypothetical faultlines within the team (Thatcher & Patel, 2012). As teams expand in size, they are more likely to have diversity among members and possible issues with coordination among members. Research has shown that individuals are more likely to find in-team members as more likeable and more trustworthy (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). Additionally, teams that are more homogenous tend to have higher commitment and cohesion, and teams that are more heterogeneous are more likely to experience conflicts (Rioedan & Shore, 1997; O’Reilly, Caldwell & Bartnett, 1989).
There are three main types of conflict experienced within teams: relationship, task, and process (Amason, 1996; Guetzkow & Gyr, 1954; Jehn, 1994; Jehn, Northcraft, & Neale, 1999). Wit, Greer, and Jehn (2012) elaborated on the three types, and defined relationship conflict as “disagreements among team members about interpersonal issues, such as personality differences or differences in norms or values” (p. 360). Task conflicts are discrepancies within the team about the specific outcomes and content of a task, and process conflicts include differences about the procedures or logistics of the particular task (de Wit, Greer, & Jehn).
Diversity in social categorization primarily leads to relationship conflict, which has been associated with reduced collaboration between members (Choi & Sy, 2010; Jehn & Mannix, 2001). Studies have also indicated that negative relationships can lead to dysfunctional and hostile behaviors (e.g., concealing significant information or sabotage, see Chiaburu & Harrison, 2008; Lyons & Scott, 2012).
However, teams are more likely to encounter status or task conflict when differences occur due to informational diversity. Task conflict seems to create more discussion within the team and leads to increases in creativity (De Dreu & West, 2001; De Dreu, 2006). Information diversity in teams provides a larger amount of knowledge resources and prior experiences to guide decisions. Research has generally supported these findings, showing that when the overlaps between relationship and task conflict are reduced, the association between performance and task conflict is indeed positive (de Wit, Greer, & Jehn, 2012).
Kooij-de Bode, van Knippenberg, and van Ginkel (2008) studied information elaboration and distribution in relation to team ethnic diversity. They argued that perceived differences led to less communication, which in term negatively influenced information elaboration and decisions. The results of their study supported these notions. Diverse teams did not elaborate on information and actually made worse decisions than homogenous teams. However, when teams were told to elaborate on the decision-relevant information there were no differences between the heterogeneous and homogenous teams.
Park and DeShon (2010) may have provided some explanation for these findings. They believed that the minority’s ability to express their own opinions was a major contributing factor to quality decision making. Most teams do not fully consider dissenting opinions or minority opinions, which is in line with the information sharing bias. They argued that goal orientation preferences might lead the team to be more open to others’ opinions. The goal orientation literature has identified learning (also referred to as mastery) and performance orientations (Button, Mathieu, & Zajac, 1996; Elliot & Dweck, 1988, Nicholls, 1984).
Learning orientations tend to be focused on the process of mastering the concepts and the information. Performance orientations are concerned with how their behaviors will be conceived and whether their performance will be viewed as competent or incompetent. Park and DeShon’s data suggested that teams with learning orientations were more likely to encourage minority opinions when compared to performance goal orientations. Politics
Individuals may contribute to the team to be truly helpful, but they may also operate under another agenda. The motives behind sharing information, generating discussion, and making team decisions may be tied back to self-serving motives and organizational politics. Witt, Hilton, and Hockwarter (2001) examined the negative effect of politics on teams. Using 172 teams, they demonstrated that perceptions of politics were negatively associated with perceived effectiveness, satisfaction, and commitment to the team. These relationships were moderated by the adoption of a team goal. Specifically, individuals who adopted the goal of the team were less likely to view politics as negative when compared to team members who had not adopted the team goal.
One of the most common biases in team research is groupthink. Groupthink was originally developed by Irving Janis in the early 1970’s. Groupthink occurs when individuals are pressured by extraneous factors to go along with a particular decision. Janis believed that groupthink was problematic for quality decisions, and the concept has been heavily popularized with mixed support. For instance, Whyte (1989) discussed how groupthink is only a matter of avoiding a greater evil due to individual aspirations, pressure to confirm to other decisions, and polarization within the team. However, some researchers believe groupthink is a valuable concept that is simply difficult to measure (e.g., Ahlfinger & Esser, 2001).
Kameda et al., (2010) evaluated the functional relations between individual contributions and team productivity. They argued that many studies worked under the pretense that every member of the team is willing to collaborate for the team enterprise (e.g, Sorkin, Hays, & West, 2001; Hastie & Kameda, 2005). Using gaming techniques, they demonstrated that it is often in an individual’s best interest to provide for the team- unless there is an abundance of individuals already contributing during which freeriding may occur. Freeriders tend to be individuals who rely on the work of others so the team will be effective (Olson, 1965). Freeriding, synonymous with social loafing, is a very popular research topic within the performance literature.
Furthermore, status within the team can affect whether deviant behaviors in the team decision making process are associated with positive or negative outcomes. Rijnbout and McKimme (2012) showed that deviance is accepted during decision making if the deviant individual holds a key position within the team. Furthermore, the deviance lead to positive outcomes and was not associated with the common negative outcomes. For instance, deviant behaviors are linked to increased creativity and innovation in teams (Nemeth, Connell, Rogers, & Brown, 2001), but also decreased team morale and satisfaction of members (Nemeth, Brown, & Rogers, 2001; De Dreu, 2006). Individuals were more willing to work with a deviant when they held a key position when compared to an individual who held an average position.
Another area where politics in teams can affect decisions is within top management teams. In 1998, Gerowitz wrote “the type and variety of backgrounds represented on a team should be related to team decision-making processes” (p. 58). He evaluated a number of demographic factors, but also leadership, perceived use of power, and political behavior. He hypothesized that the use of power or political behaviors would be associated with lower perceived adaptability, which was thought to be essential for top management teams. Although the research was conducted on a very small sample of hospital workers, the findings supported their proposition. Limitations and Future Research
One of the biggest issues within this area of research, as Gerowitz (1998) illustrates, is the accessibility to information. Studying teams in their natural settings and being able to focus on the team decision process is a very convoluted undertaking. Much of the literature in this area is done within lab settings or within a classroom. Wittenbaum et al. (2004) remarked that the hidden profile has yet to be studied in an organizational work setting and, as a result, has led to a lack of true understanding as current research omits externally imposed situations. However, even within applied settings researchers are still measuring processes that are largely cognitive and difficult to capitalize on.
Beyond these fundamental limitations, there are other measurement and research design issues. The literature is fragmented by nomological differences and mixed findings. Despite a wealth of research, there is a lack of understanding regarding individual level factors that contribute to team decision making processes. For instance, little has been done to assess personality, well-being, intent-to-leave, or psychological safety factors that may influence team decisions.
Although the current meta-analyses have done a wonderful job at review the quantitative findings, there is still a need for additional research to study the potential boundary conditions that have led to the mixed findings. In addition, there is very little research on the team decision making process over time. A potentially fruitful research endeavor would be to look at the decision making process through the entire lifespan of multiple teams using novel measurement approaches (e.g., diary reporting). It would also be advantageous to test competing models of decision making as our understanding of teams continue to grow.
Other future research areas may look at what defines a quality decision, imposed leadership, and virtual teams. A portion of prior team research has been conducted in a laboratory environment that has definitive criteria for a “good” versus a “bad” decision. The defining criteria may not be as black and white within an organizational setting. It would also be advantageous to explore how information sharing takes place when leadership roles are enforced.
Finally, an area of team’s decision making research that is relatively untapped is within the virtual team realm. Much of the current models and contentions were developed utilizing teams that met face-to-face. The decision making process may be very unique within virtual teams as the medium to communicate information is different, social category diversity may be less influential, and organizational politics may include unlike factors. Practical Implications
As the aforementioned research highlights, many team decision making processes are plagued by information sharing issues, conflict stemming from social diversity, and organizational politics. Moreover, the literature does not present a clear picture on how to best proceed in making a team decision. Sunstein and Hastie (2015) provided organizations with multiple ways to make teams smarter.
First, they believe that leaders can organize the team better by assigning specific roles and by simply letting other members speak first. Next, the members should be taught critical thinking skills and should be encouraged to challenge ideas. The team should also be rewarded so that everyone is interested in contributing. They also suggest that there should be a devil’s advocate or even an entire team dedicated at identifying the pitfalls of a possible decision. Finally, team members could attempt submitting their decisions anonymously. Conclusions
Team decision making practice and research have a current disconnect, which has perhaps led to faulty decisions. Decisions within teams are often guided by proper information sharing, ability to embrace diversity, and defer any disadvantageous organizational politics. The current research presents a number of conflicting findings, but there are specific recommendations that teams can follow. Using these recommendations will encourage better decision making and positive organizational outcomes.