Sunday, January 25, 2009

More on family size and IQ: From my recent posts on family size and IQ, you might make the mistake of concluding that you need to keep your family small in order to optimize your children's intelligence. Most of my readers are highly educated people: Let's look at IQ scores of the children of parents who attended graduate school by number of siblings. Dad first:

Mean vocab scores for children of fathers who attended grad school (N = 987)

Only child 7.91
One sibling 7.46
Two siblings 7.47
Three siblings 7.34
Four siblings 7.24
Five siblings 7.41
Six siblings 6.83
Seven siblings 7.07

Mean vocab scores for children of mothers who attended grad school (N = 493)

Only child 7.34
One sibling 7.07
Two siblings 7.46
Three siblings 7.43
Four siblings 7.11
Five siblings 6.87

The second list is shorter because not many moms with seven or more kids went to graduate school. The only-children of graduate school fathers seem smarter, but not those with highly educated moms. When families get really large, there seems to be a bit of a drop-off in IQ, but the overall picture is one of high intelligence across all family sizes.

If you want a smart child, the best thing to do--other than choosing a smart partner--is to have a bunch of kids in order to increase your odds. If you want a high average, well you've got the partner part to focus on.


  1. How about tabulating the maximum score among the children in the family? That would give some more detail on your idea that the way to have at least one smart kid is to have a lot of children.

  2. But isn't there a problem? Doesn't the GSS vocabulary score rise with age? Or is it corrected for age? If the former, then the more children, the more likely there will be an older child with a high grade. That would be a problem with the mean also.


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...