Problem solving is “what you do when you don’t know what to do.”

What a simple, straightforward definition for something often defined in much more complex ways. But problem solving doesn’t always mean the same thing. It might be the solution to a specific problem, like those that appear on math quizzes, or it might be a collection of possibilities that respond to a complex open-ended problem. But however it’s defined, problem solving is one of those skills all teachers aspire to have their students develop.

Understanding how problem-solving abilities develop is not easy, and measuring their development is even more complex. As a result, much of the research involves analysis of learners solving “knowledge-lean, closed problems that do not require any specific content knowledge to solve and that have a specific path to the answer.” (p. 866)

What this means is that “while we know a great deal about the problem-solving process in an abstract environment, we do not in fact have much insight into how students solve many types of scientific problems.” (p. 866) Not having this knowledge makes it pretty difficult to address problems that students may have as they work to solve more complex problems, like those included in an introductory chemistry course, for example.

But technology can help with the understanding of how students solve these more complex problems. The research reported in this article uses a software system that allows teachers to “track students’ movement through a problem and model their progress as they perform multiple problems.” (p. 867) The software uses case-based problems, for example, a chemistry case in which the student must identify an unknown compound based on physical and chemical tests that the student requests. There are “literally thousands” (p. 867) of possible paths that a student can take through this problem, according to the article, and the software can aggregate similar performances. Previous research has documented that the problem-solving ability of a typical student will not improve after he or she has completed about five problems using this software.

Given what other research has documented about the effectiveness of working collaboratively in groups on problems, this research team wanted to explore a “tantalizing” (p. 869) possibility: that collaborative groups might be effective in promoting the further development of problem-solving abilities. If groups were effective, would that benefit be retained when students went back to solving problems on their own? To answer those questions and another on how the nature of the group might affect the group’s effectiveness, researchers had students “stabilize” by working five problems individually; they then did five more problems collaboratively in pairs, and finally they did another five problems on their own.

Results? “Even individuals who had been given time to stabilize on a strategy adopted different strategies after solving problems in collaborative groups.” Better yet, after working with a partner, “a higher percentage of students adopted more successful strategies.” (p. 869) Based on data manipulation made possible with the software program, researchers conclude that most students improved by about 10 percent.

But that wasn’t all. Researchers grouped students according to their scores on a Group Assessment of Logical Thinking test. This instrument places students in one of three groups (based on Piaget’s theories of intellectual development) according to their level of thinking. About 50 percent of first-year college students are in the highest level. The collaborative pairs used in this research combined students from the same level for some of the pairs and students from different levels in other pairs. When a student in the lowest level was partnered with a student in the middle or high level, the lowest-level student had gains equal to those in all the other groups, “indicating that if they are paired with a student who can explain the problem and discuss it with them, they can improve their problem-solving performance significantly.” (p. 870) Interestingly, when students from the middle level were paired with those from the lower level, the middle-level students also became more proficient problem solvers.

Conclusion? “Using over 100,000 performances by 713 students on a problem, we have shown that we can improve student problem solving by having students work collaboratively in groups. These improvements are retained after grouping and provide further evidence of the positive effects of having students work in groups.” (p. 871)

Reference: Cooper, M. M., Cox Jr., C. T., Nammouz, M., Case, E., and Stevens, R. (2008). An assessment of the effect of collaborative groups on students’ problem-solving strategies and abilities. *Journal of Chemical Education*, 85 (6), 866-872.

Excerpted from Do Problem-Solving Abilities Develop in Groups?, *The Teaching Professor*, Volume 23, Number 4.