Devil's Advocates: Improving Information Sharing

Garold Stasser and his colleagues published a series of influential studies in the 1980s and 1990s regarding information sharing in groups.  Stasser's research examined groups in which members possessed a mix of shared and unshared information.  Shared information represented data which all members of the group possessed.  Unshared information proved to be unique to a particular individual.  Stasser created a series of experimental studies to examine how groups behaved in this type of shared/unshared information circumstance.  

For instance, in one study, he created a murder mystery with a series of clues.  If people put the clues together successfully, they could identify the culprit and exonerate the other suspects.  Stasser compared how individuals and groups performed when trying to solve this murder mystery.   He found that individuals actually outperformed groups.  Stasser argued that groups struggled because members tended to not share, discuss, and integrate privately held information effectively.  

Recently, Brian Waddell, Sukki Yoon, and I conducted a new study using Stasser's murder mystery.  Our research will soon appear in the academic journal, Management Decision.  Our control groups worked on the same mystery using the same instructions provided by Stasser.   In our experimental condition, we assigned one member to play the role of devil's advocate.  We asked that person to question assumptions, probe lines of thinking, and encourage people to think differently.  We compared the results of the two types of groups.  We found that groups with a devil's advocate performed much more effectively than the control groups.  Here is a chart summarizing our results:

B. Waddell, M. Roberto, and S. Yoon, (Forthcoming). "Uncovering Hidden Profiles: Advocacy in Team Decision Making." Management Decision.

Read More

Resolving Team Conflict

In Warren Bennis' terrific book, Organizing Genius, he suggests a simple method for intervening productively when conflict emerges within a team.  He learned the method from Robert Taylor at Xerox PARC.  Taylor used to refer to Class 1 and Class 2 disagreements.  In a Class 1 disagreement, the "combatants" cannot even fully explain the other sides' position.  In a Class 2 disagreement, the parties can clearly articulate the other side's views.  When a conflict emerged, Taylor would urge group members to shift toward Class 2.  In other words, he would ask them to articulate clearly and concisely the other side's point of view.  That method seemed to take the temperature down a bit, and it often helped pave the way toward a resolution.  Bennis argues that it helped the members find common ground.  Moreover, he explains that it may have used the principle of cognitive dissonance to Taylor's advantage.  In other words, the simple act of understanding and then stating the other side's position and rationale may have made opposing members more likely to embrace key elements of it. 

Read More

No Conversation Cops!

This morning, I read this interview with Blake Anderson, Adam DeVine, and Anders Holm - the creators of the show Workaholics on Comedy Central.   They describe how they brainstorm together to come up with ideas for the show.  To me, the most fascinating portion of the interview addresses the issue of how they "put the brakes on an idea."   Anders Holm explains that you don't want the same person always being the individual who is reining the group in or pruning ideas:

Adam: I don’t think any of us have a real problem, being like, “Naaaah,” to an idea. Because it’s not hurting our feelings, it’s just an idea and we’ll come up with a new one. We try not to be too precious with any of these ideas. And we’ve worked together now--like really, really worked hard together--for the last six, seven years. So we’re really comfortable knowing whether the guys are gonna love this one. Or the guys aren’t gonna love this one but I’m gonna keep pitching it anyway.

Anders: You gotta juggle that role around or else all of a sudden somebody becomes the cop. Then as soon as that person speaks it’s like, “Uhhhhggg, here comes the cop.” For us everybody puts on the badge every once in awhile.

I'm in agreement about avoiding the conversation cop phenomenon.  In working with executives, I often talk about the value of assigning someone to play the role of devil's advocate to enhance the quality of high-stakes decisions.  However, I argue that a group should rotate the responsibility for playing the devil's advocate.  Otherwise, that person can become a broken record, and other group members can begin to downplay the arguments put forth by the constant critic. 

Read More

Working on Multiple Teams at the Same Time: Positive or Negative?

Wharton Professor Martine Haas (a grad school classmate of mine) and Duke Professor Jonathan Cummings have published a new study about teams.  They collected original data within a multinational corporation about how people allocate their time among multiple teams on which they are serving.  According to Knowledge at Wharton, here is what they found:

The professors also discovered that highly skilled people were more likely to be involved in multiple teams, and that having members who were involved with multiple teams was associated with good team performance. A person with that kind of knowledge tends to be highly sought after, so his or her attention will likely be divided among many teams. And that's not necessarily a bad thing, says Haas. Teams composed of these valuable members also performed well. "That was a surprising result," she notes. "It's not necessarily causal, but it suggests if you take people who are assigned to several other teams already and bring them onto a new team, that team can benefit from their networks, knowledge and access to resources, even if they can't devote a lot of attention to the team. The exception is if the team members who are involved in lots of other projects are also very geographically dispersed -- in this case, the fact that they cannot devote much attention to the team makes the performance advantage disappear."

I find these results compelling, and they make sense intuitively.  I do wonder about one thing though.  Does it matter what TYPE OF WORK the teams are doing, or THE TIME CONSTRAINTS under which they may be operating?  Perhaps some types of work require more focused attention, while others benefit from people being able to access a wider network through service on multiple teams.  Similarly, perhaps working under a tight deadline could make it less attractive to serve on many teams simultaneously. 

Read More

Why What Others Think of You Does Matter!

Researchers Heidi Gardner and Lisa Kwan have conducted an interesting new study that explains why some teams do not perform effectively.   Their research shows that what others think of you actually does matter.  Specifically, your fellow team members' assessments of your ability have a significant effect on team performance overall.  Why?   The scholars studied the variance in team members' perceptions of each others' expertise.   Low variance means the members have similar perceptions.  High variance means the team members hold clashing perceptions of each others' expertise.  The scholars describe the high variance condition as "expertise dissensus."   They argue that expertise dissensus increases interpersonal conflict and decreases productive collaboration.  As a result, the team members experience less satisfaction and cannot work together effectively over time. 

Read More

How Teams Reject Good Advice

Prof. Mueller

Wharton management professor Jennifer Mueller and Wharton lecturer Julia Minson have published a fascinating new paper titled, "The Cost of Collaboration: Why Joint Decision-making Exacerbates Rejection of Outside Information." Minson and Mueller compared how pairs of people responded to outside input as opposed to individuals working alone.  They found that people working in pairs exhibited a greater tendency to reject outside input.   The individuals and the pairs gave initial responses to a series of questions such as, "What percentage of members of Congress are Catholic?"   Then, they had an opportunity to revise their estimate based on outside input.  As expected, the pairs demonstrated higher accuracy during their initial responses (two heads are better than one).  However, the discrepancy in accuracy disappeared after the opportunity to incorporate outside input.   Why? The individuals working alone tended to adjust their responses more so than the pairs. 

What's going on here?   A number of factors surely play a role in this phenomenon.  However, I think the general point is that teams have a tendency to be inward-focused at times.  An in-group vs. out-group dynamic emerges, whereby you exhibit an affinity for your fellow group members, and you tend to reject, marginalize, or discriminate against those in the out-group (such as the outsider providing input to the pair).  The group members also may spend time bolstering each others' confidence in the judgment at which they arrive, and that makes it difficult to alter that judgment in the future.

I recall one fascinating example of this phenomenon in action during a leadership development workshop.  My colleague Amy Edmondson was conducting a team exercise called the Electric Maze.  She invited a group of individuals on stage to work on the exercise.  After the group had a chance to plot their strategy for a few minutes, the audience members had an opportunity to offer the group advice before it started the exercise.  The group barely listened to the audience.   They had become so fixated on the strategy that they had begun to concoct that they were not receptive to outside advice.  The amazing thing is that the group had only been plotting its strategy for a few minutes when the outsiders chimed in with their input.  Yet, the group dismissed the outside input. The team already had become insular!

Read More

Team Scaffolds: Enhancing Group Effectiveness

Melissa Valentine and Amy Edmondson of Harvard Business School have published an intriguing new working paper about team effectiveness.  Valentine and Edmondson explore a concept that they call "team scaffolds" in this paper.   The scholars ask the question: "It is increasingly necessary for 24/7 shift operations to include some component of team-based work. But how can organizations support such work among constantly changing groups of people in a setting where stable teams are not feasible?"  They examine this question by studying an urban hospital's emergency room.  However, the findings apply to other settings in which people often work in unstable/transitory teams. 

The scholars begin by pointing out that many organizations cannot design and use stable teams for certain types of work.  How then can they make these teams effective, given that the team design literature suggests a certain level of structure and cohesion boosts group performance?   Valentine and Edmondson explain that scaffolding proves quite effective in the emergency room setting that they studied.  Team scaffolds consist of "a bounded role set with collective responsibility for interdependent tasks."   In other words, team roles have been so well-defined that members can step into a role and interact effectively with their peers, even though they haven't worked together as a unit on a regular basis.

The emergency room in the study underwent a major transformation.  The ER did not have a team structure prior to the change.  Patients were treated in a sequential/linear process with various functions performed by different staff members who did not work as a team.  The hospital changed its organization structure, creating what they called "pods" - a physical location with dedicated computers, counters, supplies, beds, and crash rooms. The authors explain:

The pods were each staffed by one attending, one in-charge resident and possibly another resident or intern, and typically three nurses, one of whom was designated the “Pod Lead.” The pods themselves were stable structures that persisted over time, but the staffing of the pods changed constantly. Within some five-hour periods, all of the individuals staffing the pod composition may have turned over completely as a result of shift changes staggered across roles. The nurses, residents, and attendings (collectively called providers) were assigned to a pod at the beginning of each shift. Provider pod assignments were made more-or-less at random, and a provider may have been assigned to a different pod every shift. Their “pod mates” were typically different every shift as well. Education tended to happen within each pod following the redesign, rather than through a department-wide formal rounding process. The attending and resident worked closely together as they cared for patients and informally rounded together several times during the shift, as well as during the shift change between attendings or residents. Note that the pod system connected a clearly defined set of roles (the attending, one or two residents, and three nurses) with collective responsibility for a set of tasks (the patients assigned to the pod) bounded by a shared physical location.

What happened as a result of the structural transformation?   Valentine and Edmondson found that patients’ average time in the ER fell by nearly 40% after implementation of the pods.  The key, though, is that these pods did not constitute stable teams.  However, people worked together quite effectively in these group structures.  The authors argue that the well-defined boundaries and roles facilitated effective coordination.

Read More

Who has the D? Does method encourage advocacy over inquiry at times?

Many people have embraced a terrific HBR article written a few years ago - "Who has the D? How clear decision roles enhance organizational performance". I think the article is excellent. It provides practical advice for enhancing the efficiency and effectiveness of decision-making processes. As they argue, it's so important to figure out who has the decision rights/authority before getting too deep into deliberations on a complex issue. I do think a word of caution is in order though. If we are too quick to frame a crucial meeting(s) as a decision moments, we risk having advocacy crowd out all inquiry. In other words, people may focus so much on winning the argument that the group as a whole stops learning about the problem, or about each other's perspectives and knowledge. Jumping too quickly into decision mode sometimes means that teams frame decisions as go/no go situations, rather than creating and considering multiple options. We sometimes have to remind a team that they may need to do some more collective and collaborative inquiry into the nature of the problem before shifting into individual advocacy mode. I'm not arguing for a go-slow approach. I'm simply recommending that groups remember that focusing too quickly on the "d" can lead to entrenched and polarized camps locked in a dysfunctional conversation without having shared ad integrated all the data and knowledge required to make a sound choice.

Read More

How You Select the Leader Matters

Stanford's Bob Sutton has a terrific post this week on his Work Matters blog.  He reflects on some old research regarding the selection of leaders and its impact on group performance.    S. Alexander Haslam and his colleagues published a paper in the Group Dynamics journal in 1998 called,  "Inspecting the emperor's clothes: evidence that random selection of leaders can enhance group performance."  They examined groups performing an experiential exercise called the survival game.  Many MBA students have participated in this exercise during introductory organizational behavior courses.  In this Haslam study, the researchers compared group performance under four conditions:

1.  Formal selection of a leader (self-ratings by group members)
2. Information selection of a leader (group members picked a leader through a discussion)
3. Random selection of a leader.
4. No leader selected.

The findings demonstrated that the highest performing groups had used the random leader selection process.  Naturally, I don't think we want to recommend random selection in real organizations based on the results of this study.  However, the research does have important implications regarding the impact that the leader selection process can have on group dynamics.  Sutton writes,

"I especially like how it implies that just THE PROCESS of selecting the leader can provoke group dynamics that undermine the performance of the group as a whole.  That is worth considerable attention as this is something that selection committees and such often forget -- and consistent with findings from many corners of the behavioral sciences that show 'what you do is as important as how you do it.'"  

Read More