Teams Blog

People in my line of work don’t want to admit there is little objective evidence team building works. Reports that it does come from people like me, and the people who pay people like me, all of whom have a vested interest in it working. Few team builders bother to create metrics proving a measurable change occurred, as I like to, and fewer customers are willing to pay for tracking those metrics. It has bothered me that I could not point to a study done by scientists with nothing to gain showing for sure that team development impacts team performance. Then I cam across a promising study title from last year: “Does Team Building Work?”

An army of scientists from the Univ. of Central Florida and the U.S. Army Research Institute said in their article in Small Group Research that recent study results were mixed at best, with many studies showing no impact. So they searched for every study of team building interventions that measured performance from a 47-year period. Using data from the 103 studies they found, the researchers state that teams which had performed team-building activities also had better processes, and their members felt better about their teams.  Performance was stronger, too, but the correlation was not as high. You can have better satisfaction and processes without having better measurable performance, obviously. Team building had a weak impact on the team’s thinking processes (cognition). The overall impact of team building was much stronger for teams with ten or more members, probably because they had more problems and thus more room for improvement. For the specific numbers and source information, see my study summary.

Team building aimed at setting team goals and clarifying roles was better than activities that tried to improve interpersonal relations or problem-solving. This is consistent with my observations, though I am biased. My services focus on the first two; teach problem-solving only in context of solving the team’s task issues; and only target interpersonal problems directly as a last resort. Fixing group dynamics eliminates most person-to-person issues while providing more bang for the team’s bucks.

All of that is good news for anyone tying to convince someone to do team building. But don’t get too excited. There are many limitations to this encouraging study. As best I can tell after rereading the same paragraphs five times, it does not differentiate between the games- and ropes-courses method of team building and approaches like mine that help team members agree on formal structures and processes. This disappointed me, since it won’t help prospects choose between these approaches.

The study does not prove that the team building activities caused the positive outcomes. In theory, it could be that teams with high performance are more likely to do team building, perhaps because their efficiency gives them more time for it. Or other factors might encourage both better team performance and more team building, and thus be the root causes. Based on my experience, I think that is part of the story, but those factors are not enough to create high-performance teamwork.

Then there’s the problem of the “intervention effect,” the teamwork equivalent of the placebo effect in medicine. Drug studies can’t just compare people who took a drug with those who didn’t. They have to compare those who took the drug with, for example, people who took a pill that looked like the drug but was made of plain sugar (a “placebo”). Usually the people given placebos do better than those who took nothing. If the drug takers did no better than the placebo takers, it’s back to the lab for the drug maker. By the same token, there is some evidence that doing anything for a team makes the team perform better for a while, perhaps because the employees are happy just to be noticed. Have everybody stand on their heads for a minute and they might report higher morale (until they get back to work).

Nonetheless, all of you with managers or employees (or clients) who refuse to do team building now have some solid ammunition from nearly 50 years of research. In my experience, resistors have experienced something like a team rafting trip that brought no lasting results back at work. Or they were forced to take personality tests that were helpful but only addressed a tiny part of the overall problems, problems that remained unaddressed. Team building works, the UCF research team says. You simply have to choose the kinds that directly target the problems you are facing.

Action Item: If you have been thinking about doing some team building, contact TeamTrainers for a diagnosis of the issues prompting the thoughts. If we can’t help you directly, we’ll know someone who can.

Researchers have been looking into manufacturing, supply chain, and project management for years, often making recommendations to managers. Maybe you have followed their advice on the job. But there’s a problem, according to professors Francesca Gino and Gary Pisano: “Most formal analytical models in operations management (OM) assume that the agents who participate in operating systems or processes—as decision makers, problem solvers, implementers, workers, or customers—are either fully rational or can be induced to behave rationally.” In other words, scientific theories about how to run a plant or project assume that people:

• identify and react only to relevant information;
• have the same preferences in every situation;
• consider all options before making a decision; and
• make those decisions without emotion.

Oh, yeah, sounds like every workplace I’ve been in.

Gino, of the Univ. of North Carolina-Chapel Hill but headed to Harvard Univ., and Pisano of Harvard, point out in a 2008 paper that people effects on a  system’s performance have permeated other fields of research, from economics and accounting to the law. But “a ‘behavioral perspective’ has largely been absent in the field of operations,” they write. Because of this, current OM models can’t really explain the difference between one firm’s performance and another’s. In turn, managers don’t find OM theories very useful. Especially relevant to Teams Blog is the authors’ point that “behavioral operations” researchers needs to look into how individual cognition and “social norms and systems affect operations.” Gino and Pisano say based on previous research (cited below) that this will lead to very different predictions about what will fix specific issues.

In an interview I asked Gino, an assistant professor of organizational behavior, why OM scientists resist research on the effect of human behavior. She said there is “skepticism from some people that maybe it is not dramatic or very significant…” She noted that her co-author’s interest was sparked by going into organizations and seeing what worked, which suggests that other academics have not done that. However, she said, “The researchers, the more they hear, the more they understand that it is important to study the psychology of people.”

People effects have explained results that defied scientific theories in other fields. In the OM world, this could explain “the tendency of projects to run late and over budget or the tendency of organizations to over commit their R&D resources,” the article says. Researchers have identified many biases and questionable rules-of-thumb that affect our decision-making. Gino and Pisano provide a somewhat depressing list of 19 shortcuts humans take in their decision-making that can mess up the results. Some include:

• “Information avoidance—People’s tendency to avoid information that might cause mental discomfort…”
• “Confirmation bias—People’s tendency to seek information consistent with their own views or hypotheses”
• “Law of small numbers—People’s tendency to consider small samples as representative of the (entire) populations from which they are drawn”
• “Sunk costs fallacy—People’s tendency to pay attention to information about costs that have already been incurred and that cannot be recovered… when making current decisions”
• “Conservatism—People’s failure to update their opinions or beliefs when they receive new information…”
• “Hindsight bias—People’s tendency to think of events that have occurred as more predictable than they in fact were before they took place”

Take another example the professors explore, the “anchoring and adjustment” bias. People often start their thinking from a particular point, sometimes without a good reason for it, and then stay too close to that point. In one study, software developers given a higher anchor to start with ended up with higher final estimates than when they were given lower or no “anchors,” the article says. Sales forecasts are often off because they start with the previous year’s sales instead of an unbiased analysis of this year’s market.

Of course, behavioral operations researchers and managers can’t erase human bias. However, Gino and Pisano write, “operating systems can be designed in such a way that systematic errors are eliminated, or at least their negative consequences reduced.”

I asked Gino what advice she would give, for instance, a chief operations officer whose IT planner tends to anchor too closely to industry averages. “First, you need to be aware of the bias, which is a very simple lesson, but it is hard to recognize,” she said. Have someone act as a devil’s advocate, she suggested, asking the planner to bring alternatives to the table and questions like, “How did you come up with this number?” I would add, based on something else she said, that you cannot push the person for a certain number and be surprised when it turns out wrong. In the IT scenario, don’t anchor yourself to industry averages if the planner offers good reasons not to.

While other scientists are catching up to Gino, Pisano and other… okay, I can’t resist calling them “BO researchers*” at least once… take a look at the biases table in the article and maybe you’ll find your own answers. Or call me and I’ll show you how to account for people effects in your operation.

Sources:

• Bendoly, E. (06), K. Donohue, and K. Schultz (06), “Behavior in Operations Management: Assessing Recent Findings and Revisiting Old Assumptions,” Journal of Operations Management 24(6):737.
• Boudreau, J., W. Hopp, J. McClain, and L.J. Thomas (03), “On the Interface Between Operations and Human Resources Management,” Manufacturing & Service Operations Management 5(3):179.
• Gino, F., and G. Pisano (08), “Toward a Theory of Behavioral Operations,” Manufacturing & Service Operations Management 10(4):676.

*BO is American slang for a person’s smell or “body odor.”

How team members “think, make decisions, plan, design, perceive, and remember”—in other words, how information is processed as a group—will clearly impact how the team performs.  But many managers focus on the outputs of the process instead of the process itself. For example, if they don’t like a plan, they will send it back with some changes and questions. When they have to go through this several times on every plan, they understandably become frustrated. Often, though, they will either yell at the team for whatever they perceive the problem to be or start doing the plans themselves. Instead, perhaps they should stop to look at the understanding the team has about what they are planning or the planning process itself.

It may have been Edwards Deming who first said the majority of what appear to be individual performance problems are actually process problems (I confess I didn’t record the source). The AntiClue blog has made the same point, and it’s one of performance consultant Don Clark’s main causes of performance problems (3/4 of which are mostly out of the individual’s control). Almost all process problems boil down to someone not having the information they needed when they needed it, if you include as “information” knowledge, skills, procedures, external decisions, etc. A study from last year confirms that managers need to pay much more attention to how their teams get, use, and store knowledge. Psychology professor Leslie DeChurch and management professor Jessica Mesmer-Magnus teamed up to draw relevant data from 65 studies into team cognition. The American Heritage Science Dictionary defines cognition as, “The mental process of knowing, including awareness, perception, reasoning, and judgment.” The professors describe it as including both “the nature of team member interaction” and “conditions that dynamically enable and underlie effective teamwork,” so it’s both a process and a group mental state. Psychologists Nancy Cooke and Jamie Gorman, the sources of the first quote in this post, put it succinctly: “team mind.”

Cognition is measured two ways. One compares the understanding of some topic across each pair of members on the team. For example, the similarity and accuracy between members A and B, B and C, and A and C are averaged to get a team cognition figure. The other way looks at the team’s overall view of some topic, measuring members A, B, and C together. The DeChurch/Mesmer-Magnus study looked at the ways these two approaches; the type of team (decision-making, action, and project teams); and other possible factors might affect the relationship between cognition and team performance. I won’t go into all the details, but it was clear that the more a team had a shared perception of its work, the better it performed. Though more true on subjective tests like ratings by managers, it also seemed to impact measurable performance. This appears to operate in part by improving work processes and member motivation, but a higher level of team mind by itself raised performance.

The authors suggest that managers pay attention to whether perceptions are shared among team members and in what ways. Cooke and Gorman use the example of a team that works with boilers, pumps, electronics, and motors. Everyone on the team could see a strong connection between all four of these in their work. In a more complementary kind of shared cognition, a couple of members might see connections between the first three items, while another set of members might only see a connection between pumps and motors. Between all the members, however, the team mind sees a connection of all four. In the latter scenario, you can surely see the potential for misunderstanding if those different views were not understood and respected. That is, complementary cognition is okay so long as the people who don’t see the pump/motor connection defer to those who do on matters affecting the motors.

Accuracy of cognition is also important. For example, say you work in a pharmaceutical plant where following a standard operating procedure (SOP) is required by government regulation. It’s not enough for one or two members to understand the SOP thoroughly. If the whole team doesn’t understand it, you increase the risk of a regulatory breach.

DeChurch and Mesmer-Magnus make the interesting suggestion of performing job task analyses on the work the team does and then “structure important support systems (e.g., measurement, performance appraisal, and reward structures) to develop and shape the collective cognition needed for successful teamwork.” I was trained on job task analysis while working at Los Alamos. An outsider observes people while on the job to determine the steps they go through and thus the knowledge, skills, and information necessary to succeed at the job. That done, the researchers say, leadership and training would be likely methods to ensure the team develops the proper cognition.

Or “team mind,” as I think I’ll start calling it. That term makes for all kinds of punny opportunities:

Q. What kind of cognition does a group of soil scientists need?
A. A dirty mind.

Sources:

1. Cooke, N. J., & Gorman, J. C. (in press), “Assessment of Team Cognition.” In P. Karwowski (Ed.), International Encyclopedia of Ergonomics and Human Factors, 2nd Ed. Taylor & Francis Ltd.
2. DeChurch, L., and J. Mesmer-Magnus (2010), “The Cognitive Underpinnings of Effective Teamwork: A Meta-Analysis,” Journal of Applied Psychology 95(1):32.