Friday, June 13, 2008

IQ, race, and college graduation: In my earlier post, I showed that the mean IQ for white, 4-year college grads dropped 9 points in about 4 decades to roughly 105. But what about large minority groups like blacks and Mexican Americans? The General Social Survey doesn't have enough cases to examine 10-year changes for these two groups, but the analysis of whites shows that the low average IQs have been seen over the last 20 years, so let's look at averages for blacks and Mex-Ams since 1990:


Mean IQs of 4-year college grads--1990-2006 average

Blacks 100.85 (N = 87)
Mexican Americans 99.52 (N = 27)

So minority students are completing college even though they have a 5 point IQ deficit when compared to whites. How are they able to graduate with less intelligence? More perseverance? Graduating, but with lower GPAs? Faculty allowing them to skate? The data don't allow us to say.

One other interesting question is differential graduation rates of black males and females. In my own classes, black girls outnumber the guys 4 to 1, easy. But according to the GSS, graduation is 2 to 1, so of course we'll go with that. Do more women attend and finish because of a higher mean IQ? No. Using the same years as described above, I looked the mean IQ of all blacks: women have a mean of 91.67, while the male mean is 93.32. (Remember that these estimates of IQ are vocabulary questions, and the black-white gap is not as large as with overall IQ tests).

So why do black women do better and go further with the same IQs? I suspect that the pathological aspect of black culture puts more pressure on black guys to not conform to what is perceived as the white system. You're not a real man, but kind of a bitch to do what everyone tells you to do. Women don't feel this pressure as much, and can conform and still feel like a woman.

Another possibility is patience and mental self-discipline. School requires a lot of self-restraint, and many black men, given a choice of whether to attend school or not, might not want to put themselves through more of what was very difficult for the past 12 years.

4 comments:

Anonymous said...

Underclass black men go for brass ring work that will not pan out for the vast majority (athletics, crime, entertainment) or do nothing.

Anonymous said...

Do white males have the same graduation rates as white females?

Anonymous said...


Do white males have the same graduation rates as white females?


No, but females go in for the soft subjects that are not much use in the real world.

Carl said...

In the spirit of La Griffe du Lion, graduating from college can be considered like an IQ test. I know there are other factors, but we can gain some insight into the observations of the post by ignoring those. Due to the difference in IQ distributions between whites and blacks (I'm ignoring the Mexican Americans because their IQs are not as well characterized and for simplicity), even if there is a common IQ threshold to graduate the mean IQ of the graduates will be substantially different. To quantify this I did some quick calculations using MATLAB. I used a white IQ distribution of normal with mean 100 and sigma 15, and black with mean 85 and sigma 15. If I apply a minimum IQ to graduate from college of 90 for both whites and blacks, then the average IQ of the graduates is white=106.2 and black=100.4. This is virtually identical to the reported values. This doesn't exclude the possibility that minority students are graduated undeservedly, but it certainly casts some doubt on that idea.

I found it interesting that to get the reported mean IQs I had to set the graduation IQ threshold so low. I thought it might be interesting to perform a similar analysis for the earlier post on the drop in average IQ over the decades. I'm going to assume that the reported IQ values are unaffected by the Flynn effect, since presumably they were normalized within their own time frame. Using the same style of analysis as above but excluding the black portion I get the following approximate IQ threshold to graduate as a function of time:

1960s - 103
1970s - 98
1980s - 94
1990s - 86
2000s - 88

Also note that while about ~41% of the white population had the brainpower to graduate in the 60s, this increases to ~79% in the 2000s. So we see a direct correlation between reduced standards and more people capable of college graduation.

Please note that I realize I'm excluding many things from the above analysis and it is in no way intended to be definitive. It does, however, provide some insights with a dirt simple model and a few minutes of calculation.