The Girl Next Door (are sex and studying substitutes?)

by on March 18, 2014 at 8:27 am in Education, Uncategorized | Permalink

That is a new piece (pdf) by Andrew J. Hill, at the University of South Carolina.  It seems your child should beware the girl next door:

Parents are concerned about the influence of friends during adolescence. Using the gender composition of schoolmates in an individual’s close neighborhood as an instrument for the gender composition of an individual’s self-reported friendship network, this paper finds that the share of opposite gender friends has a sizable negative eff ect on high school GPA. The eff ect is found across all subjects for students over the age of sixteen, but is limited to mathematics and science for younger students. Self-reported difficulties getting along with the teacher and paying attention in class are important mechanisms through which the e ffect operates.

For the pointer I thank the ever-excellent Kevin Lewis.

1 Z March 18, 2014 at 8:36 am

So the girl next store causes Johnny to have bad grammar, unable to manage pronoun-antecedent agreement? That would explain why so few American kids take a foreign language. All of the masculine and feminine nouns must be baffling.

2 Marie March 18, 2014 at 8:52 am

Gordon Neufeld years ago made good points about peer orientation and schooling. One of them was that in the early grades, being peer oriented helped schools, because that’s where the peers were, so the kids wanted to be there. But once the kids get to high school (earlier, probably, now) they go to school (if they must) to hang out with peers but they are able to circumvent teachers so that they can spend their school time focusing on their friends. So they are physically in the school, but the peer orientation works against the educational goals of the school.

3 dearieme March 18, 2014 at 9:17 am

Youngsters with no intellectual interests prefer wasting time with their chums to wasting time on lessons of no concern to them. Could be.

4 Michael G. Heller March 18, 2014 at 9:25 am

The article sounds boring. But the good blogpost title reminds me of Berkeley California in late 1960s when kids could still play safely on the street and get to know the girl next door. I’m sure I tried hard to impress her at John Muir School. It was ‘elementary’ before ‘high’.

5 C.B. Reed March 18, 2014 at 3:38 pm

“The article sounds boring”?

Sigh. Go back to

6 DougT March 18, 2014 at 9:30 am

Yet another reason to homeschool or use single-gender schools. I wonder how long the effect persists?

7 Z March 18, 2014 at 10:00 am

Single-sex education has the biggest impact on boys. That should come as no surprise to anyone familiar with biology and humans. It is also why we will never see public schools going to single-sex classes, much less single sex schools.

As far as whether it lasts, it depends upon what you are measuring. IQ is a constant, but learning relative to IQ is not. Smart boys will learn more in a single-sex high school than smart boys in a public gladiator academy. Seeds among the stones and all that.

8 GiT March 18, 2014 at 1:34 pm

Whatever IQ is, it isn’t a constant.

9 Y March 18, 2014 at 6:15 pm

Quiet you, Z wants to get his racism and sexism on. Your inconvenient science is getting in the way of his bias dressed up in science-y language.

10 X March 19, 2014 at 12:40 am

We obviously need an analogue of Godwin’s law, with ‘racism’ and ‘sexism’ substituting for ‘Hitler’. As for Z’s statement about IQ being a constant, he formulated it badly. Adult IQ has indeed been shown to be very insensitive to any interventions in childhood. Look up some twin adoption studies.

11 ZZ March 19, 2014 at 5:16 pm

Such a study would require a meaningful definition of IQ, no?

12 JWatts March 18, 2014 at 10:08 pm

“Whatever IQ is, it isn’t a constant.”

Argument from ignorance?

13 david March 18, 2014 at 10:16 am

Single gender schools have been noted to delay marriage significantly, which sharply reduces fertility, so the selective use of single gender schools by the academically-oriented may have interesting long-term effects.

14 X March 19, 2014 at 12:43 am

Students bound for good colleges shan’t marry young anyway. OTOH the potential to reduce teen pregnancy and single motherhood may be interesting. Ya think?

15 Daniel March 18, 2014 at 11:48 am

Americans – declaring war on sex (among many other things) sincer … forever.

16 X March 19, 2014 at 12:43 am

It’s called civilization.

17 Daniel March 19, 2014 at 11:23 am

Can’t tell if you’re being sarcastic or stupid.

18 The Other Jim March 18, 2014 at 9:35 am

>the share of opposite gender friends

Those with many opposite gender friends are what we call “attractive people.”

Being attractive correlates with lower intelligence, yes, we knew that. I would suggest that it’s because they have more opportunities to do fun things that are not studying, and they have a greater chance to get what they want out of life without studying.

19 david March 18, 2014 at 9:42 am

the OLS actually points in the opposite direction – attractiveness and GPA go hand in hand, so your remark on correlation is flatly wrong. the authors speculate parental sponsorship driving both attractiveness and high performance

it is the IV which the authors try to use to remove parental involvement.

20 Z March 18, 2014 at 10:02 am

Perhaps The Other Jim was trying to tell us he is is not good looking?

21 Chris S March 18, 2014 at 11:13 am

As a high school slacker with a 3.0GPA and 99th percentile on standardized tests, GPA does not always correlate well with intelligence. (I am not sure standardized tests correlate as much with intelligence as test-taking skill, but that’s another debate.)

(My preference was goofing off with my computer, trying to make free long distance phone calls to connect to far away BBSes, not girls, which made me uncool then but, now that nerds are ruling the world, is paying dividends.)

My guess is that GPA and attractiveness ARE correlated because instructors like and are charmed by attractive people more – especially in post-pubescent high school – so grade them higher in subjective areas. I was (am?) pretty nerdy looking so had neither that effect or completed homework to help my grades. Fortunately I am very good at taking tests.

22 david March 18, 2014 at 11:19 am

this would work. It would also work with the IV results, since being resident near youths of the opposite gender plausibly does not cause attractiveness

23 matthew March 18, 2014 at 10:45 am

They use gender composition in their neighborhood as an instrument for number of opposite gender friends. So their instrumental variables strategy is impervious to the bias you mention, unless you’re claiming all the dumb attractive kids happen to live in neighborhoods with lots of opposite gender kids.

24 Doug March 18, 2014 at 11:18 am

Don’t know how strong of an affect this might be. But consider people choosing housing will be influenced by friends, family and community members being near by. One major source of friends as an adult are the parents of the friends of your kids, who are usually the same gender. Hence families with better community ties and more friends are more likely to live near the friends of their kids. This means that these families will probably exhibit a same-child gender bias when looking at their neighborhood. Needless to say families with better community ties probably have smarter kids.

Again the magnitude of this effect might be trivial. But you can’t necessarily discount it right away.

25 david March 18, 2014 at 11:26 am

the authors deny it:

Essentially, parents do not choose the locations of their homes based on the gender of school-going neighbors. The parent component of the survey used by this paper includes a question in which the parent is asked their motivation for their residential locational choice, and there is no evidence of neighborhood gender composition playing a role. … Potential correlation with observables is investigated in the online appendix by performing balance tests in which the instrument is regressed on a set of individual and background characteristics; these are shown to have no systematic e ect

26 Doug March 18, 2014 at 11:49 am

Appendix Figure 8 in the paper shows “Close Friends (Parents)” as the second most important factor when choosing housing. The authors showed that gender composition isn’t a direct factor, but they fail to anticipate how other housing factors may bias neighborhood gender composition.

Like I said many adult parents are friends with the parents of their kids friends. And most kids friends are same gendered. Hence picking neighborhoods with close friends should bias the neighborhood gender composition. The authors failed to anticipate this, but they could easily check the magnitude of the effect by checking the correlation between neighborhood gender composition and those who chose housing based on “Close Friends”.

27 david March 18, 2014 at 12:26 pm

Is Close Friends (Parents) indicate friends of the parents, or friends of the kid?

28 lump1 March 19, 2014 at 9:40 am

Well that’s a weird selection bias. The neighborhood with the greater number of opposite gender friends is almost inevitably going to be the neighborhood that is *denser* (read: more urban). So maybe the measured effect is simply that less dense neighborhoods correlate with better grades. Then it wouldn’t be about sex at all, but about how property sizes correlate with grades, and nobody is going to be surprised by that.

29 Daniel March 18, 2014 at 11:47 am

Being attractive correlates with lower intelligence, yes, we knew that.

No, it doesn’t. Being attractive correlates with intelligence, health & living a long life. The real world isn’t a Hollywood movie.

30 Marie March 18, 2014 at 4:09 pm

@The Other Jim,

I thought those with many opposite gender friends were generally what we call “gay”?

31 JWatts March 18, 2014 at 10:10 pm

“Being attractive correlates with lower intelligence, yes, we knew that. ”

I thought the evidence points the other way?

32 Someone from the other side March 19, 2014 at 2:40 am

Last I looked into it it was a some sort of a bell curve like thingy but the mode was definitely shifted towards the right vs the IQ curve.

33 Rahul March 18, 2014 at 10:05 am

Is the noted effect sex-asymmetric? Was wondering because of the way Tyler phrased his comment.

34 david March 18, 2014 at 10:19 am

No; the authors report that the effect is significantly larger for girls. That is, the girl next door should presumably beware your boy even more 😉

35 Peter March 18, 2014 at 10:17 am

So you’re telling me high school students are distracted by members of the opposite sex?
I’m shocked, shocked I say!

36 X March 19, 2014 at 12:44 am


37 john personna March 18, 2014 at 10:24 am
38 tt March 18, 2014 at 7:46 pm

i think thats true everywhere

39 JKB March 18, 2014 at 10:29 am

Wait a minute. Is this just a heterosexual phenomena? If the solution is same-gender schools will that only concentrate the distraction on same-gender attracted students?

Perhaps the distractions are more refined as I discovered long ago. Mixed-gender instruction wasn’t to terribly distracting until in my sophomore English class some of the girls who had blossomed over the summer wore their 9th grade sweaters.

Perhaps is the school day involved more than just a hour or two of actual instruction, the distractions of comely classmates would be less eye-catching. In the battle between boredom and boobies, there is no contest.

40 X March 19, 2014 at 12:46 am

Perhaps is the school day involved more than just a hour or two of actual instruction, the distractions of comely classmates would be less eye-catching. In the battle between boredom and boobies, there is no contest.

This. Single-sex classes are still better, but this.

41 Nikki March 18, 2014 at 10:34 am
42 delurking March 18, 2014 at 10:46 am

I’m sorry, this is backwards. Getting good grades and getting along with the teacher, especially in math or science, reduces the likelihood of an adolescent having opposite-gender friends.

43 matthew March 18, 2014 at 10:57 am

Again, their instrumental variables strategy is impervious to this bias.

44 david March 18, 2014 at 11:06 am

according to the OLS, at a correlational level getting good grades is associated with having opposite-gender friends, so bias issues aside, delurking’s remark is still wrong. As the authors suggest, getting good grades is plausibly caused by parental involvement, which then causes coeducational extracurriculars and social events

45 matthew March 18, 2014 at 10:56 am

It’s unclear whether the “distraction” is merely the presence of members of the opposite sex, rather than the higher probability of romantic involvement with one of them. Or it could be something else entirely that has nothing to do with sexuality. Would be much more informative if they had a comparison between gay and straight people in their analysis.

46 Doug March 18, 2014 at 11:11 am

Even better would be if they interact the result with digit ratio.

47 DougT March 18, 2014 at 12:21 pm


48 Daniel March 18, 2014 at 11:45 am

Ah yes, another justification for sexual repression – it makes boys get good grades.

What kind of an autistic eunuch do you have to be to see that as a good thing ?

49 Autistic Eunuch March 18, 2014 at 2:53 pm

They told me if I didn’t eat the marshmallow now, I would get 2 marshmallows later.

50 dangerman March 19, 2014 at 11:08 am

^ I lol-ed.

51 whatsthat March 18, 2014 at 12:05 pm

too bad about Stacy, but what about Stacy’s mom?

52 Norman Pfyster March 18, 2014 at 12:20 pm

I hear she’s got it going on.

53 Nathan W March 18, 2014 at 12:15 pm

Here’s a classic one, regarding influence of friends during adolescence

If anyone ever demonstrates any signs of going through challenges, parents move to ostracize the child who is skipping classes, smoking, or what have you.

Then, some years later, when all children of “good parents” have completed their “good work” to “protect children”, the rebels are essentially left with no connections to the “good students”, “good parents”, or anyone else who could provide them with connections.

Know any stoners who got decent jobs? Respect them. They got there on their own. They must be very smart indeed, and their work must be second to none, because when it comes to the stoner, the minor flaw in the “good student” easily becomes (legally permitted) cause for immediate dismissal.

Know any dealers who struggle to find decent work? Ask if they would stop selling if they could find better work, then help them find it.

When it comes to sex … well, I think everyone likes sex. It’s hard to ostracize people for liking sex. And the kids like sex so much that if you try to break them from their partners, parents may see themselves faced with more real estate than they need. It’s so hard to convince people that sex is bad that even after years of religious and moral indoctrination, kids still love it.

54 Micah March 19, 2014 at 12:42 am

Thanks for contributing the lemming/Reddit/doesn’t understand people take on this subject.

55 Viet March 18, 2014 at 2:01 pm

I wonder what would the effect be for LGBTQ+ populations?

56 Engineer March 18, 2014 at 4:34 pm

Single-sex schools are absolutely the way to go. As a bonus you often get a non-nihilist administration and teaching staff.

57 Marie March 18, 2014 at 4:52 pm

I always think of school uniforms as useful signalling, I’d guess the same is partially true of single sex schools. Anyone who dares that would dare other useful things.

58 X March 19, 2014 at 12:48 am

This. Also, school uniforms are cute.

59 Thor March 19, 2014 at 12:57 am

Define “dare”! And define “useful” in the phrase “other useful things”!

60 X March 19, 2014 at 2:38 am

Aspies’ sandbox is over there →

61 M March 18, 2014 at 5:21 pm

Self-reported difficulties getting along withthe teacher and paying attention in class are important mechanisms through which the effect operates.

Seems like kids who are more focused on appealing to and connecting with their peers (or at least their opposite sex peers) are less focused on pleasing authority figures and listening to them.

Not a problem for those kids with genuine, even omnivorous intellectual interest in school but let’s be real – most kids want to please the teacher and please their parents and don’t want to look bad in front of other kids.

Having supportive peer motivations could generally erode all those motivations more than it enhances them.

62 YetAnotherTom March 18, 2014 at 6:25 pm

It’s not cool to get caught trying in high school. That’s one way good grades negatively affect ones dating life. For males, Aspergian tendencies are more common for the very smart ones I’d bet. That lack of social thinking can’t help ones dating life either.

63 Micah March 19, 2014 at 12:39 am

At the risk of sounding Straussian, I always find the Rorscach comment threads to be the most interesting.

64 ezra abrams March 19, 2014 at 9:51 pm

Familiarity breeds attempt

65 SumDumGuy March 21, 2014 at 9:35 am

This was a classic Seinfeld episode…

66 Amber Bracelet Outlet Canada April 2, 2014 at 11:37 pm

Comments on this entry are closed.

Previous post:

Next post: