Brilliant: The Science of Smart

Can Tough Competition Hinder Academic Performance?

A new book on the science of winning and losing explores the benefits of competition--and the dangers of pushing students too far

  • Share
  • Read Later
Doug Corrance / Getty Images

Top Dog, a new book by Po Bronson and Ashley Merryman about “the science of winning and losing” is in large part a celebration of competition. The authors of the bestselling NurtureShock explore the benefits of what they call “competitive fire” — stories of Olympic swimmers, champion chess players, and upstart political candidates who reached the top by racing someone else. But just as interesting are the cases in which we do better without the element of competition. Sometimes, it turns out, competing against others can actually make our performance worse.

Bronson and Merryman describe an experiment in which researchers gave 124 Princeton University underclassmen a test that drew its questions from the GRE, the graduate school admissions test. For some of the students, the investigators added to the stress of this difficult exam in two ways. First, the students were asked to report which high school they’d attended and how many of their high school classmates were also at Princeton. “This was intended to make most test-takers feel as if they were alone at Princeton, that they were lucky to be at Princeton, and that they had barely made the bar for admittance,” Bronson and Merryman explain.

(MORE: Relax, It’s Only a Test)

Second, researchers further added to students’ stress by labeling the test as an “Intellectual Ability Questionnaire.” Bronson and Merryman again: “They wanted the test’s title to be threatening to the students, to make the students fear that, if they did poorly, the test would reveal they lacked the true ability to be at Princeton.” The other group of students answered the questions about high school only after taking the test, when it could no longer affect their performance, and their exam went by the less-threatening name “Intellectual Challenge Questionnaire.”

The results? Students in the first group answered 72% of the questions correctly; those in the second group got 90% of their answers right. By subtly manipulating the competitive stress felt by the participants, Bronson and Merryman note, the researchers “were able to engineer an 18% difference in their test scores.”

(MORE: Highlighting Is a Waste of Time: The Best and Worst Learning Techniques)

The research literature is full of such findings. Take a classic study of the phenomenon known as “stereotype threat,” the apprehension felt by members of certain groups, such as female or African-American students, that a poor performance will confirm a negative stereotype about them–“girls aren’t good at math” or “blacks aren’t college material”. Published in the Journal of Personality and Social Psychology in 1995, the study had groups of students, both white and black, take the same test; as in the Princeton experiment described by Bronson and Merryman, the test was presented slightly differently to each group. Some participants were told that that the test would evaluate their intellectual abilities; others were told that it was a laboratory puzzle task that did not assess ability.

Black students who thought their intelligence was being evaluated did worse on the test than their white counterparts, while black students who believed they were simply figuring out puzzles (a condition the researchers termed “stereotype-safe”) did much better, equaling the white students’ scores. A follow-up study found that self-doubts and negative racial stereotypes were more likely to intrude on the thoughts of African-American students who anticipated taking an evaluative test; African-American students who expected to complete a puzzle exercise were less likely to encounter such thoughts.

So when does competition help our performance, and when does it hurt? One more story from Bronson and Merryman helps make the distinction clear. In the mid-1990s, they recount, the commandant of the U.S. Air Force Academy became concerned that the number of cadets experiencing academic trouble was on the rise. Economists Scott Carrell and James West studied the cadets and noticed a pattern: cadets with lower grades improved academically if they spent time with cadet friends who did well in school. So the academy deliberately engineered the composition of the squadrons of entering cadets, grouping cadets with lower GPAs and SAT scores with cadets who’d achieved high grades and scores.

(MORE: How to Use Technology to Make You Smarter)

The intervention was a failure. The low-performing cadets actually did worse than before. Why? Unable to compete with their high-flying squadron mates, the low-performers gave up trying. Bronson and Merryman deliver the takeaway: “Contests only work when it’s an even match up, or a close race, such that the extra effort becomes the decider between wining and losing. People need at least a fighting chance. When leaders are not challenged, they coast a little. Those too far behind stop trying as hard, lacking any sense that winning is feasible.”

Human beings are supremely sensitive to context, to the cues we sense in our surroundings, and never more so than when we’re performing. When you feel stressed or threatened, you can try mentally reframing the situation as a game or a challenge; when young people feel anxious, parents and teachers can help by downplaying the evaluative nature of the event. But when we feel strong and capable, when we feel like a contender — then we can use the spur of competition to reach new heights.

MORE: Does College Put Kids on a ‘Party Pathway’?


you don't need a study to know some people perform better under stress, some perform worse under stress. That's why not everyone can be the gold medalist in Olympia game. Performing under stress in itself is a quality. No all individuals have it. Certain job requires people to perform under stress. Not everyone can make it. I don't know your point in this article.


“Contests only work when it’s an even match up, or a close race, such that the extra effort becomes the decider between wining and losing."  People may not like this, but maybe we should split classrooms up into the advanced, average, and behind levels again.  I know you think little Johnnie is an allstar, but he needs to compete with kids at his level.  It doesn't do anyone any good to group everyone in one large group and teach to the lowest level.  It just teaches them its ok to underpreform


What does it mean that gym is overwhelmingly focused on competitive games?


This study has nothing to do with competition.  Its nothing more than testing anxiety and negative reinforcement.  That doesn't work.  Try pitting the students against each other and offering rewards for good performance instead of negative reinforcement.  This everyone is a winner no matter what mentality is creating a bunch of entitled children who don't want to work for anything if its "too hard".

CharlieBrown 1 Like

There is no revelation here.  The research on frustrating a learner to the point that motivation is lost and performance/achievement declines is relatively old.

It is a problem that so much in education has no hard research basis.  Much of what passes for educational research in this country is essentially "opinion survey" based.  And as if that weren't enough of a deterrent to effective practices, hard research is often ignored.  A prime example of this is the "high stakes" state testing of mandated objectives conducted in many public schools.  Multiple independent studies have decried the impact of these tests on the quality of teaching and learning as well as their extremely limited correlation with excessive litanies of state objectives that (1) often cannot be tested and (2) require more instructional time and resources than are ever provided.


As per my observational experience, while in competition the participant's real self's natural efficiency gets enforced to be  hidden for a while and  so he / she just fights aggressively just to win in the competition beyond his / her capacity of natural performance , hence his / her brand of real performances get harassed. in an unjust manner.

     - A.R.Shams's Reflection  -  Press & Online Publications


Forgive me for being slow on this, but "testing anxiety" is a given in any (and, arguably) every testing situation.  You're always going to have those students who feel more or less of that anxiety than they care to admit.  Whether that anxiety impairs performance is debatable.

With that being said, what exactly was the point of this article.  To me, it seemed to 1) re-state the obvious about such anxiety, and 2) offer suggestions (that are not currently implemented) to reduce that anxiety in a controlled setting.  

Sounds like a lot of speculation to me.


"Human beings are supremely sensitive to context" is the most important finding. Athletes train for competition so that when they perform they are not distracted and can deliver full potential. Athletic coaching also works to facilitate success by confidence. Athletics-to-academics there is almost no correspondence in context.