Friday, June 20, 2008

Predictors of celibacy: As a follow-up to the previous post, I looked for General Social Survey variables that predict that a young person (aged 18-25) is celibate. Below I show trimmed models where independent variables with effects not even close to significance have been removed. These include: size of place, race, athleticism, father's level of education, political orientation, and trust of other people.


Logit coefficients--males, N = 67

Church attendance .265*
IQ .053*
Income -.405*
How healthy -.510
Constant .453

Pseudo-R-squared .373

*p < .10, one-tail test


For young men, the following predict being celibate: frequent church attendance; a high IQ; and less income.

Now for the girls:


Logit coefficients--females, N = 56

Church attendance .282*
IQ .032
Income -.094
How healthy -1.260*
Constant 2.767

Pseudo-R-squared .170

p < .05, one-tail test


Going to church a lot and not being healthy predict that a girl is refraining from sex. More income makes young men more appealing, and high intelligence is probably associated with a lack of the interpersonal skills needed to win over girls, but income and IQ don't matter if you're a girl.

4 comments:

  1. Anonymous7:08 PM

    Jason Malloy had a great post on sex and intelligence
    on GNXP awhile back.

    B.B

    ReplyDelete
  2. It's odd that poor health is a greater predictor of celibacy in males than in females. After all, sex is more physically demanding from the man's part.

    ReplyDelete
  3. How healthy = how pretty

    Maybe that's part of why youngsters are more celibate these days -- the girls are tubbing out so much that no one wants them.

    ReplyDelete
  4. You should check if there's a nonmonotonic relationship with IQ. A higher IQ increases social skills up to about 120 or so...which means 90% of the population. Your coefficient should swing the other way if you're expecting a social skills effect.

    ReplyDelete

Meta-analysis of clinical trials: Eat walnuts

I am always looking for easy eating choices that are good for you. This new meta-analysis of 26 clinical trials looked to see if walnuts ma...