Question: In a study on determinants of children born in the Philippines,22 the following results were obtained:
The variables are:
ED = years of schooling of the woman
LWH = natural logarithm of present value of husband's earnings at marriage
AMAR = age of the woman at marriage
SURV = survival probability at age 5 in the province
RURAL = residence in rural area (dummy variable); this variable is supposed to capture search and schooling costs
Age = age of the woman
The explained variable is number of children born.
(a) Do the coefficients have the signs you would expect?
(b) Using the t-ratio for AMAR and the R2 for the two equations, can you tell how many observations were used in the estimation?
(c) Looking at the t-ratio of AMAR, can you predict the signs of the coefficients of the other variables if AMAR is deleted from the equation?
(d) Given that the dependent variable is number of children born, do you think the assumptions of the least squares model are satisfied?