We often perceive competition as something bad in education. I promise this blog post will not be a plea to simply reintroduce it. But I’m bringing this topic up because a new study by Corwin Senko and Sarah Pallone suggests that perhaps we should not only look at if students want to perform better than others, but also why they want to do so. Because that can vary…

Senko and Pallone base their study on the goal complex model. This model distinguishes between the goal someone pursues and the reason they pursue it. To make this more concrete, imagine you have two students who both want to be the best in their class. At first sight, they have the same goal: to perform better than others. But perhaps student A does this because he or she simply enjoys measuring themselves against other people. Or maybe he just likes a challenge. Maybe he simply finds it personally important to do his best or to perform well. Student B may pursue exactly the same goal for very different reasons. For example, because he feels pressured, is afraid of disappointing others, or wants to prove himself. Or simply because he wants to get rewarded.

The researchers describe how student A has more autonomous reasons for competing, while student B has more controlling reasons. Previous research already suggested that this distinction could be important. Performance goals seem to be associated with less desirable outcomes. Especially when people pursue them for controlling reasons. But much of that earlier research had quite a few limitations. It was often based on questionnaires. You can, for example, find that students who experience a lot of pressure are also more anxious, but that doesn’t tell you whether that pressure causes the anxiety. Yes, the difference between correlation and causation again.

Senko and Pallone therefore wanted to take this a step further and conducted two experiments. In their first experiment, 261 American university students took part. They first learned how to solve Sudoku puzzles. Next, they received a series of new puzzles. The researchers divided the students randomly into three groups. One group was simply asked to solve the puzzles. A second group received the goal of outperforming the other participants, presented in a controlling manner. Therefore, the researchers used phrases such as you must and you should and emphasised that they had chosen the goal. More than that, the researchers also told the participants that their performance would be evaluated afterwards.

The third group also received the task to perform better than others. So the competition did not disappear. But… this time the researchers put much more emphasis on autonomous motivation. The researchers used more choice-supportive language. They also acknowledged, for example, that participants might not like competing with others at all and emphasised that they could decide for themselves whether they wanted to pursue that goal. The feedback the students received afterwards was also not presented as an evaluation by others. No, it was information participants could use to review their own performance. Almost all participants in the experiment, by the way, agreed to pursue the competitive goal.

So what did the researchers find in this experiment? Participants who received a performance goal framed in a controlling way experienced the task more as a threat and less as a positive challenge than those who received no performance goal at all. With the autonomously framed competition, the authors couldn’t find this difference. These students responded roughly like the control group.

That feeling of threat, in turn, was linked to more anxiety while solving the puzzles, and greater anxiety was linked to poorer performance. The researchers thus found an indirect path from the controlling performance goal, through threat and anxiety, to lower performance. They found no such path for the autonomous performance goal. An important difference.

But there was an issue with this first experiment. The researchers had wanted to use easy and difficult Sudoku puzzles. The problem, however, was that the easy puzzles turned out not to be that easy at all. The participants solved an average of 67 per cent of them correctly, compared with 58 per cent of the more difficult puzzles. So in fact, there were mainly difficult and even more difficult tasks. Because of this, the researchers could not properly test an important hypothesis. Perhaps controlling competition becomes a problem especially when people are confronted with something genuinely difficult.

That is why they conducted a second experiment, this time with 249 other students and a different kind of task. This time they used matrix reasoning problems. Here, you see a pattern in which one element is missing and you have to work out from several possibilities which element completes the pattern. Personally, it mainly reminds me of what you sometimes see in IQ tests.

This kind of task allowed the researchers to vary the difficulty much more effectively. And it worked: participants solved, on average, 95 per cent of the easy problems correctly, 85 per cent of the medium problems and 61 per cent of the difficult ones. Useful.

The researchers kept the rest of the design the same. And again, they found the same initial effect. Participants who were told in a controlling way to perform better than others experienced more threat. Those in the autonomous condition did not. But this time, the researchers could also examine what happened as the problems became more difficult.

With the easy problems, it hardly made any difference. The control group solved an average of 94 per cent correctly, compared with 96 per cent in both competitive conditions. With the medium problems, the differences were also not significant. But with the most difficult problems, there was a difference. The control group scored an average of 65 per cent, the group with the controlling performance goal 60 per cent and the group with the autonomous performance goal 62 per cent. In the analyses, the controlling group performed significantly worse than the control group; for the autonomous group, the difference was not significant.

That last point is important to read carefully. This study does not show that competition improved performance. The autonomous competition group did not perform better than the group that had received no competitive goal at all. What these two experiments do suggest is that it is too simple to say that trying to perform better than others is harmful in itself.

There is a tendency to discuss education in simple oppositions: competition versus cooperation, performance goals versus mastery goals, intrinsic versus extrinsic motivation, and many other topics that regularly appear on this blog. This study suggests once again that one of those oppositions is more complicated. Two students can both want to perform better than the rest, but the psychological meaning of this can be quite different.

I have taken quite some time in this post to explain the research, so that you also understand that you can’t simply translate this one-to-one into everyday classroom practice. These were not experiments in real classrooms, but relatively short studies with American university students who had to solve Sudoku puzzles and matrix problems. The researchers did not just manipulate the competitive goal itself. They also changed how they presented that goal: the language they used, the amount of choice participants experienced, and how they framed later feedback and evaluation. Based on this study, you can hardly conclude that there is no harm in introducing rankings and competitions in every classroom.

Because, and this is important, the results are much more convincing as evidence for the disadvantages of controlling competition than for the benefits of autonomous competition. So the conclusion is rather that a student who enjoys measuring themselves against a classmate does not necessarily learn less because of it. A little competition or the desire to be the best for once does not automatically have to be pedagogically suspect. But once wanting to win turns into having to win, the story seems to change.