Do We Learn Best Collaboratively or Individually?

I read an interesting article about collaboration and worked examples today. Worked examples, for those not in the know, are teaching objects that explicitly show students the steps for how to solve a particular type of problem, such as the one below for how to add fractions:

Worked examples are an effective form of instruction. When learners are given worked examples and asked to study them, they learn from them. Worked examples are also very efficient. Learners who study worked examples tend to report lower feelings of mental effort while learning significantly more than if they had tried to discover the solutions on their own. Often when you hear people talk about the effectiveness and efficiency of explicit or direct instruction, they will mention the research on the worked example effect.

The world hasn’t learned everything there is to know about worked examples. Which is superior, studying an already worked-out example on a piece of paper or having a teacher reproduce the worked example on the board while thinking aloud? What’s better, studying a worked example first, or doing some exploratory or goal-free activities before being shown a worked example? My own PhD research intends to investigate and expand upon emerging research that indicates that students may not be very good at choosing to use worked examples during learning, but would instead prefer the inefficient route of mucking about with problems on their own.

This brings me to the article I read, Can collaborative learning improve the effectiveness of worked examples in learning mathematics? (Retnowati et al., 2017), in which the authors explored an interesting question: Do learners learn best when studying worked examples individually or in groups? One might predict that working together with other learners could increase the effectiveness of worked examples, as group members could build off each other’s understanding while picking apart the steps of the worked examples together. Another way to look at it would be that working in groups can be quite distracting, and that collaborating with other novices introduces irrelevant information that gets in the way of learning how to solve the problem as an expert would.

What did the researchers find? In the first experiment of the study, it appears that individual study of worked examples won compared to collaborative study for high-complexity problems, and that collaborative study was sometimes more effective for low-complexity problems. The cognitive load measures were all over the place. The results were reported in Table 1:

In the second experiment, similar results were obtained as in the first experiment, with the researchers reporting that “no overall support was found for Hypothesis 1 that students would benefit from studying collaboratively rather than individually when using worked examples. Instead, the reverse result was obtained, with individual study superior to collaborative study on both similar and transfer tests for high complexity problems” (Retnowati et al., 2017, pg. 673). Experiment 2 also expanded upon the findings of Experiment 1 by testing whether students who engaged in unguided problem solving collaboratively would outperform those who engaged in unguided problem solving individually. Interestingly, if not predictably, they did, as shown in the bottom half of Table 2:

It’s by reading articles like this that I strengthen my understanding of how to help students learn. Providing students with examples is effective, withholding examples from students when they could benefit from them is bizarre and irresponsible, and if teachers want students to have a shot at making sense of the examples they’re provided, it’s probably best that they study them individually, away from the attentional competition and extraneous noise that is the group dynamic. While the results of this one study seem to imply that teachers use collaboration with caution, we probably shouldn’t do away with it entirely. Students do gain some domain knowledge while hunched over a worked example in groups (especially, it seems, when the information is less complex), and collaboration may be an effective way to practice collaboration. Collaborative problem solving may also be an option whenever students must solve problems for which no worked example or teacher explanation exists. Needless to say, folks, I would be wary of any self-proclaimed expert, “thought leader,” or “education guru” who claims that collaborative learning is a vastly superior method to learning something individually without any evidence to back up their claims. There’s clearly more to the story.

Zach Groshell @mrzachg

Reference

Retnowati, E., Ayres, P., & Sweller, J. (2017). Can collaborative learning improve the effectiveness of worked examples in learning mathematics? Journal of Educational Psychology, 109(5), 666–679. https://doi.org/10.1037/edu0000167