LINDA WERTHEIMER, HOST:
Parents and educators have long assumed that peers matter. If you are at a high school or college where you are surrounded by serious students, you're more likely to take your studies seriously. If your friends are party animals, you're more likely to want to party, too.
NPR's social science correspondent Shankar Vedantam, who joins us regularly on this program, recently heard about an unusual social engineering experiment that tried to apply what's known about peer effects to the real world.
Shankar, walk us through what happened.
SHANKAR VEDANTAM, BYLINE: Well, the story unfolds at the U.S. Air Force Academy, Linda. This is in Colorado Springs. It's one of the military's elite training schools, very tough to get in, and graduates typically go on to join the officer corps in the military. It's a very tough course, mentally, physically, emotionally demanding. And administrators there noticed a couple of things.
One: Some of the cadets were dropping out. And often the cadets who dropped out were the ones who were weakest academically, especially when it came to test scores showing verbal ability.
The second thing they noticed is that when weaker students happened to be in squadrons with stronger students, the weaker students started doing better. So being around the better students made them better as well.
I spoke with Bruce Sacerdote at Dartmouth College, along with his colleagues Scott Carrell and James West, and administrators at the Air Force Academy. Sacerdote told me they hit upon an idea.
BRUCE SACERDOTE: Hey, maybe what we could do is rearrange these squadrons to provide help for the least academically able of those cadets. We could stick them in a squadron with lots of high ability cadets and that's going to reduce dropout and raise GPAs and so forth.
WERTHEIMER: That seems to makes sense. They were finding that when cadets with better test scores happened to be in the same squadrons as cadets with weaker scores, the stronger students had a positive effect on the weaker students. So presumably you could set that up deliberately.
VEDANTAM: Exactly. Now, the cadets who were the average performers, Linda, the ones with the test scores in the middle, they didn't seem to play a role in this. So for two years the researchers pulled out these middling students into their own squadrons, and they stuck the students with the lowest performance into the same squadrons as the students with the best performance.
WERTHEIMER: How did that work? Did the weakest students do better?
VEDANTAM: They did not, Linda. In fact, the news is even worse than that. When the students who were the weakest academically were stuck in the same squadrons as the top performers, they actually did worse. Here's Sacerdote again.
SACERDOTE: The results are pretty clear that the least able students had lower GPAs than low ability students in the control group. As soon as we realized that we were harming the least able students, we stopped the experiment.
WERTHEIMER: So what was going on? It seems to make sense that the least able students would benefit from being around the most able students.
VEDANTAM: It did make sense, Linda. And it points to why common sense can sometimes be misleading. It turns out that in the squadrons that had the strongest and the weakest students together, these students were so different from one another that they actually splintered into two different groups. They were in the same squadron but they were in different social universes. The good students hung out with other good students. The weak students hung out with other weak students. And now pure effects(ph) worked in the way that you didn't intend; the weak students were surrounded by other weak students, and their performance actually went down.
It turns out that the group Sacerdote thought was unimportant was actually crucial. It was the group of middle ability students in regular classes - these students acted like the social glue between the best students and the worst students. And they may be what allowed the best students to actually influence the weaker students. When you took them out, you took the glue out of the system.
WERTHEIMER: What happened to the middle group? How did they do once they were cut out of the pack?
VEDANTAM: Well, the middle group started to do better, Linda. When they didn't have the influence of either the better students or the worst students, the middle students by themselves, their test scores went up. And for the policymaker, this presents a dilemma, because it suggests that in regular squadrons where everyone is mixed together, the performance of the middle students is actually adversely affected. You pull them out, they start to do better. But when you pull out, the weakest students start to do worse.
WERTHEIMER: Thanks, Shankar.
VEDANTAM: Thank you, Linda.
WERTHEIMER: Shankar Vedantam regularly joins us to talk about social science research. You can follow him on Twitter @HiddenBrain. You can also follow this program @MORNING EDITION.
(SOUNDBITE OF MUSIC)
WERTHEIMER: Thanks for checking in with us this morning. You can stay with NPR News throughout the day. ALL THINGS CONSIDERED will pick up where we leave off this afternoon. Transcript provided by NPR, Copyright NPR.