Group consultation has long been lauded as the best process for problem solving in organizations because it results in a wider range of solutions than most individuals can design on their own. Now there's a new study, from psychologist Patrick Laughlin and his colleagues at the University of Illinois, that shows that the approaches and outcomes of cooperating groups are not just better than those of the average group member, but are better than even the group's best problem solver functioning alone.
These findings underscore the importance of communication in problem solving and have important implications for managers and anyone else who works as part of a team. Far too often, a leader—who, by virtue of greater experience or wisdom or skill, is deemed the ablest problem solver in a group—fails to ask for input from team members. Equally dangerous, members of a team often relinquish problem-solving responsibilities to the leader and fail to provide her with important information for moving forward on a decision.
The consequences of this vicious circle? Suboptimal solutions, bad choices, wrong directions, and avoidable errors.
Don't go it alone
Laughlin's data tells us why even the best problem solver operating individually will be beaten to a demonstrably correct solution by a cooperating unit.
|If you're the brightest person in the room, you're in trouble.|
— James Watson,|
Nobel Prize winner
First, the lone problem solver can't match the diversity of knowledge and perspectives of a multiperson unit that includes him. Other members will have had experiences with similar or related problems that will allow the team to recognize fruitful versus fruitless choices more clearly and quickly. Furthermore, this diversity of input can do more than merely add to the storehouse of information that the best problem solver can employ; it can also stimulate thinking processes that would not have developed in wholly internal monologues. We all can recall being led to a productive insight by the comment of a colleague who didn't deliver the insight itself but who sparked an association that did the trick.
Second, the solution seeker who goes it alone loses a significant advantage—the power of parallel processing. Whereas a cooperating unit can distribute the many subtasks of a problem-solving campaign among its members, the lone operator must perform each sequentially. This requirement considerably extends the time spent on the effort. In addition, it strains the capacities and energies of the problem solver because the subtasks often include activities that are daunting in their difficulty (e.g., information integration), time-consuming in their execution (e.g., library/Internet research), and demotivating in their tediousness (e.g., fact checking).
The Nobel Prize-losing error
These findings echo a remarkable interview published last year on the fiftieth anniversary of the publication of perhaps the most important scientific discovery of our time—that of the double-helix structure of DNA, as revealed in the Nobel Prize-winning work of James Watson and Francis Crick. The interview, with Watson, was designed to inquire into those aspects of the duo's efforts that had led them to solve the problem ahead of an array of highly accomplished rival investigators.
At first, Watson ticked off a set of contributory factors that were unsurprising: He and Crick had identified the problem as the most important one to attack. They were passionate about it, devoting themselves single-mindedly to the task. They were willing to try approaches that came from outside their areas of familiarity.
Then he added a stunning reason for their success: he and Crick had cracked the elusive code of DNA because they weren't the most intelligent of the scientists pursuing the answer. According to Watson, the smartest of the lot was Rosalind Franklin, a brilliant British scientist who was working in Paris at the time.
"Rosalind was so intelligent," observed Watson, "that she rarely sought advice. If you're the brightest person in the room, you're in trouble." That comment illuminates a familiar error seen in the actions of many well-intentioned leaders.
Another type of error stems from a failure to collaborate. It's called captainitis, and it refers not to the tendency of a leader to assume all problem-solving responsibilities but to the equally regrettable tendency of team members to opt out of responsibilities that are properly theirs.
The error gets its name from the sometimes-deadly type of passivity exhibited by crew members of multipiloted aircraft when the flight captain makes a clearly wrong-headed decision. Accident investigators have repeatedly registered disastrous instances when even an obvious error made by a captain was not corrected by other crew members.
Consider the following exchange, recorded on an airliner's flight recorder minutes before it crashed into the Potomac River near Washington National Airport in 1982:
Copilot: Let's check the ice on those tops [wings] again since we've been sitting here awhile.
Captain: No. I think we get to go in a minute.
Copilot: [Referring to an instrument reading] That doesn't seem right, does it? Uh, that's not right.
Captain: Yes, it is.…
Copilot: Ah, maybe it is.
[Sound of plane straining unsuccessfully to gain altitude]
Copilot: Larry, we're going down!
Captain: I know it.
[Sound of impact that killed the captain, copilot, and seventy-six others.]
Captainitis is not limited to air travel. In one study, researchers tested the willingness of well-trained nurses to give up their decision-relevant responsibilities regarding a patient once the "boss" of the case—the attending physician—had spoken. To perform the experiment, one of the researchers made a call to twenty-two separate nurses' stations on various surgical, medical, pediatric, and psychiatric wards. He identified himself as a hospital physician and directed the answering nurse to give 20 milligrams of the drug Astrogen to a specific ward patient. In 95 percent of the instances, the nurse went straight to the ward medicine cabinet, secured the ordered dosage of the drug, and started for the patient's room to administer it—even though the drug had not been cleared for hospital use, the prescribed dosage was twice the maximum daily dose set by the manufacturer, and the directive was given by a man the nurse had never met or even talked with before on the phone.
|The key to decision-making success is for the leader to avoid engaging alone in the processes that lead up to the final verdict.|
In drawing conclusions from their results, the authors of the hospital study made a telling point. They concluded that in fully staffed medical units like the ones they examined, it is natural to assume that multiple "professional intelligences"—i.e., the doctors', nurses', and assistants'—are working to ensure that the best decisions are made. But in fact, under the conditions of the study, only one of those intelligences—the physicians'—may be functioning.
It appears that the nurses unhooked their considerable professional intelligences in deferring to the doctor. Yet the nurses' actions are understandable. Regarding such matters, the attending physician is both in authority and an authority.
That is, the doctor is, first of all, in charge and therefore able to sanction noncompliant staffers. Second, the doctor possesses the superior medical training that can lead others to defer automatically to his or her expert status.
Accordingly, we shouldn't be surprised when medical staffers are reluctant to challenge a physician's treatment pronouncements. Nonetheless, we should be more than a little disquieted by this behavior, not just because of the way it could play out during our next hospital visit, but because of the way it could affect any work setting, including our own.
Implications for leaders
What common lesson flows from the two kinds of errors we have considered? Leaders attacking a knotty problem that possesses an objectively correct solution must collaborate unfailingly with team members toward its resolution—even when they are the best informed or most experienced or ablest of the group. This means setting up systems that ensure collaborative exchanges whether or not the collaboration seems necessary. To do less is a fool's gamble.
But isn't there a different type of gamble that a fully collaborative leader takes? Doesn't this approach risk the notoriously poor outcomes of decision by committee? No.
The recommendation here isn't to employ vote taking or nose counting when making hard business determinations. In fact, the recommendation here isn't for joint decisions at all in such instances. The final decision is properly the leader's alone to make. That's one thing leaders are paid for, typically because they've given evidence of being able to make such choices better than the people who haven't achieved leader status.
However, the key to decision-making success is for the leader to avoid engaging alone in the processes that lead up to the final verdict. It is these predecisional processes that, when jointly undertaken, will benefit the sole decision maker so richly.
If leaders who arrange for regular team input can expect to achieve problem-solving gains, might they also expect to lose something else in the bargain—for instance, subsequent rapport with and input from those whose ideas are rejected? Sometimes members' egos can be bruised and they can feel discouraged if the leader doesn't adopt their proposal or favored course of action.
Fortunately, when inviting cooperative efforts, leaders can take an approach that can generate high levels of collaboration while avoiding this potential problem. From the outset, leaders need to encourage everyone with a stake in the decision process to make a contribution to it and, simultaneously, to assure all concerned that each contribution—while perhaps not the deciding factor—will be a factor in the process.
Thus the leader must make the commitment that, even though many recommendations may not be followed, each is important to optimal decision development and will be given full attention.
That may not sound like much of a commitment, but when properly implemented, it's more than enough.