Is Problem Solving a Team Sport or an Individual Effort?

Have you ever wondered, “Is this particular problem one that a team should tackle, or […]

Have you ever wondered, “Is this particular problem one that a team should tackle, or one that an individual should handle?” A Knowledge@Wharton article shared the results of an experiment attempting to answer this question. According to the paper, Task Complexity Moderates Group Synergy, the reason for asking whether teams or individuals are more efficient at getting work done was that “Scientists and managers alike have been preoccupied with the question of whether and, if so, under what conditions groups of interacting problem solvers outperform autonomous individuals.”

This is an interesting claim. I would like to see the data that underpins it. Who is having trouble deciding whether a problem-solving task is fit for a team or for individuals? In my nearly 30 years in the field of business problem solving and continuous improvement, I have never faced this question from a customer or colleague. True, many managers are guilty of assigning problems of all sorts to an engineer or expert, underutilizing the talent of the entire workforce. But in my experience, humans seem to have at least a basic intuitive sense of when one human is suitable for a task, and multiple humans will be required. This is anecdotal and not empirical evidence.

In any case, the research team led by Wharton professor of operations, information, and decisions Duncan Watts found that simple tasks are best accomplished by individuals, while difficult ones are more efficiently completed by a group. This conclusion has the ring of truth to it. It also reminded me of a certain Theory on Brontosauruses.

The Experiment

According to Prof. Watts, “A manager is kind of stuck a little bit because they don’t really know how to evaluate the complexity of the task that they’re looking at.” The research team set up their experiments using a class of tasks known as constraint satisfaction and optimization problems, or CSOPs. This allowed them to vary task complexity in a systematic and principled way. Specifically, experiment participants were asked to solve the real-world problem of assigning students to dorm rooms. This began as a simple task and grew more complex as researchers added constraints such as more students, fewer rooms, and students who must or must not reside in the same room.

According to the article, this approach was unique. It allowed researchers to manipulate the complexity of the same task, rather than simply giving the participants different kinds of tasks, as most previous studies did. While ingenious, a question remains unanswered. How does a manager who is stuck on how to evaluate task complexity get unstuck?

What Is Complexity?

The researchers seem satisfied with their definition of complexity, using CSOP for the purpose of running their experiment. I’m not so sure we can do justice to complexity with a simple, one-size-fits-all definition. What do we mean by task difficulty or complexity? Something with multiple parts or elements? Large, simple things can have multiple parts, so not that. Having multiple options or variables? Possibly, but as with CSOP it’s probably not how many but more a question of how these variables interact. How do we take into account the subjective experience? An observer’s familiarity with the situation? What might look complex to a novice may look simple to someone who has repetition and practice in facing similar scenarios.

Most of us would say that complexity is something that makes our heads hurt. We can’t immediately wrap our minds around it. In other words, a complex problem is one that’s beyond the capability of one individual to easily handle. However, this is highly subjective to the individual. Even if we controlled for a class of problems and had a set of managers at similar levels of capability, there is a problem. In the context of this experiment, defining complexity as something that’s hard for one person to figure out on their own leads to circular reasoning. Unless we can quantify domain or task-specific complexity, this question becomes one more of philosophy than of science.

How Lean Thinkers Decide Who Should Solve a Problem

Working in teams may be all the rage in business today. The message of the paper seems to be that managers shouldn’t assume teams are always superior to individuals in problem solving. In lean thinking circles, we call this the “don’t be the hammer looking for the nail” principle. We shouldn’t jump to apply a proven solution every time without first understanding the context and situation. I think we can say the same for the findings of this research paper. Well-led teams are more efficient than individuals at solving CSOPs, and to a degree, complex problems in general. But I think managers should rely on better and more practical heuristics based on the objective problem and subjective experience of its complexity.

How do lean thinkers decide whether a problem-solving task is better suited for an individual or a team? Through decades of experience, there are some guidelines we follow. We rely on team-based problem-solving when redesigning a shared work area, a connected series of processes, or any domain that requires cross-functional insights. The kaizen event is an example of a format born from this type of need. On the other extreme, we ask individuals to generate small, simple, local improvement ideas. These should be based on things that bother them about their own work and are largely within their power to change. There are various creative idea suggestion schemes that support such efforts. In both cases, teams or individuals can often arrive at solutions through observation or trial-and-error.

When an organization needs to optimize the overall flow across a value stream, it’s often obvious that this goes beyond the capability, capacity or interest of an individual. Value stream mapping and transformation projects are team-based. On the other hand, an individual is often well-suited to lead even a complex project to improve one performance area of a value stream by applying the DMAIC six sigma approach. Quality circles and TQM teams can fall somewhere between these two, depending on their selected theme. As processes become less directly observable, individuals and teams rely more on data to arrive at solutions.

Many Hands Make Light Work

It never hurts to have experiment-based, empirical evidence to answer a question like, “Do I assign this problem to a team or to an individual?” But there is some irony here. We have a team of researchers setting up a complex task (an experiment) to demonstrate that teams are better at doing complex things, while individuals are better at doing simple things. Was this really a complex problem that required a team approach? It seems like an individual observer with a few ounces of common sense and basic observational skills could have answered this question more efficiently.

It may benefit a manager puzzled by how to assign such tasks to talk to their fellow lean thinker, study these methods and form a heuristic. One person is enough for a small, simple task. A team is needed for larger, more complex tasks. Most cultures pass down packets of reusable knowledge to the next generation in the form of axioms, proverbs, and adages. Many hands make light work. Too many cooks spoil the broth. Proverbs run in pairs. Common sense is not so common. And so forth.


The original article can be found at: Gemba Academy