<p>It’s not quite the same effect. To do well on the SAT, you aren’t in direct competition with other students. Their gain doesn’t mean your loss, and your gain doesn’t mean their loss. Theoretically, every student in the room could score a 2400 regardless of the performance of those around them.</p>
<p>If you’re sitting in a room with 24 other students taking the SAT, you have no idea whether they are superstars or strugglers (particularly if you don’t know anyone else in the room, which is frequently the case). The sound of other anonymous students scratching answers is not the same as seeing Tiger Woods show up in your golf tournament. A better equivalent would be if the top-scoring kid in your program who had already gotten a 2400 once showed up in the room for some reason, but even then, that kid’s getting another 2400 doesn’t really diminish YOUR chances of getting one.</p>
<p>However, there is a related phenomenon, in which average kids stay average because their expectations, and the expectations for them from everyone around them, are average. A B student expects to get Bs and therefore gets Bs, because their teachers expend B energy on them, they do B work, and their parents lavish B praise. Likewise, an A+ student gets A+ attention and A+ expectations from parents. Self-efficacy plays a huge role in standardized test scores; it’s one of the reason I hated hearing my SAT students (back when I tutored) say “I’m just not good at taking tests.” That kind of attitude sabotages you.</p>
<p>“In my school, it’s the girls who are math people.”
“For the girls and math thing, at my school there are more girls taking Calculus BC than guys. But then again the youngest person in Calculus AB is a guy.”</p>
<p>Doesn’t matter what the reality is. The persistent stereotype in society is that women are innately worse at math than men, for whatever reason, and when that stereotype is made available and accessible to female test-takers they do worse - AND the male test-takers do better.</p>
<p>Interestingly enough, when you take away the stereotype and state that there are no gender differences, not only do the women perform better but the men perform worse, which evidences the other side of the superstar effect - that if you believe you are a superstar (even erroneously) you will do better on average despite your abilities. The same was true on race - when black and white students took an exam and were simply told that there were no racial differences on the exam, the black students did better and the white students did worse than in the condition in which they were NOT told that and were asked to identify their race before hand. Thus the students in the “no differences” condition scored about equally.</p>
<p>Claude Steele, the father of stereotype threat research, is the recently appointed provost at my university and he’s given several interesting lectures on this. My own research is also about the effects of racial disparities, stereotypes, and prejudices, except in health and health education instead.</p>