Do the less intelligent have more children?

In my post below responding to David Goldstein’s implication that intelligence has been subject to strong directional selection I put up a chart from the GSS which illustrated the fact that women who have weaker vocabularies tend to have more children. If you don’t believe that intelligence “means anything,” and that it isn’t heritable, read no further.
On the other hand, if you think intelligence as measured by something like a vocabulary test is important and that it is heritable, below the fold I’ve placed some charts which show the same relationship between number of children and vocab score as the more general data set (though all are limited to women). Again, on the X axis you have the CHILDS variable from the GSS, that is, 0 to 8 or more children. On the Y axis you have WORDSUM, which shows the mean number correct for each class of women with a particular number of children. I have filtered by variables to show how the trends play out. I’ve also dropped WORDSUM mean scores where the N for the category was less than 10. I do caution not to over-read every twist and turn of these trends too seriously because of small N’s, especially as the number of children increases (e.g., women in graduate school have a very mean low fertility which drops off very quickly).

Highest level of education attained

Race

Political views

Religious views

Observation: I think the educational data tell us quite a bit and suggest that socialization and expectations are very important, more important than the lack of foresight one assumes is more common among the less intelligent. A quick regression suggests that this is so. The beta for WORDSUM is around 0 and not statistically significant when DEGREE is included as another independent variable when CHILDS is the dependent that these two are predicting (albeit, they still don’t predict much of the variation in CHILDS).
Methodological note: Remember that the data plotted above are the mean WORDSUM scores within each category of children. The number of individuals within each category doesn’t show up here.