Recall in the last few exercises, we looked at the relationship
between the number of births a woman had in her lifetime, and education.
Separately, we looked at the relationship between the number of births and the
age at first sex. Now, let's include both of these independent variables in a
multiple regression model. Make sure that the variable
nobirth is clean: that is, it only captures information for women.
Restrict the sample over which you will run the regression, to the female
population who are at least 40 years of age.
reg nobirth sexage
educ if age>=40&gender==2
Source | SS df MS Number of obs = 287
-------------+------------------------------ F( 2, 284) = 22.06
Model | 214.989524 2 107.494762 Prob > F = 0.0000
Residual | 1383.81187 284 4.87257701 R-squared = 0.1345
-------------+------------------------------ Adj R-squared = 0.1284
Total | 1598.80139 286 5.59021466 Root MSE = 2.2074
------------------------------------------------------------------------------
nobirth | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
sexage | -.1381859 .0473065 -2.92 0.004 -.2313017 -.04507
educ | -.1995234 .0380789 -5.24 0.000 -.274476 -.1245707
_cons | 9.258195 .8917576 10.38 0.000 7.502902 11.01349
------------------------------------------------------------------------------
In this regression, higher levels of education seem to reduce the number of births, conditional on age. This effect is statistically significant. Recall that this result is for the sample of women over 40 years of age. Previously, when we were not controlling for age of first sex, we found that education had a very similar effect on total number of births for women over the age of 40, in sign and magnitude.