I continue to be amazed at the predictive power of age of first intercourse for so many other things, like strip club attendance, sexually transmitted disease, number of partners, cheating in marriage, etc. However I'm quite sure it is not the age of first intercourse which causes these other things, but rather it is correlated with the true unobserved variable that does cause the other thing. Call it sensation seeking...or whatever you want, but it is clear there are large differences in teenagers in terms of these variables. Above you find the percentage of teenagers at Wellsley that are still virgins by major. Clearly there is some sorting going on, maybe partially driven by IQ. The chart comes from this excellent post.
And this post follows on with still more interesting data. Finally this could also say something about risky behavior among teens.