# a) Find the smallest and largest values of children in the sample.
# What is the average of children? 
# Does any woman have exactly the average number of children?  - NO
summary children
Mean                         2.2678


# b) Do the same for the variable educ.
summary educ
# ? There are more less educated women or more educated women?
# mean < median

# c) What percentage of woman have electricity in the home? - 14.02%
smpl electric ==1 --restrict
611/4361 = 
smpl --full

freq electric

# d) Compute the average of children for those with electricity 
# and do the same for those without. 
# Comment on what you find.

summary children --by=electric


# e) Suppose you want to observe how the education affects the fertility rate.
# Specify the equation you want to estimate and 
# use the data to estimate the model using OLS.  - children = educ + constant
# Interpret the coefficients and test a hypothesis whether educ has any effect on fertility rate.

ols children 0 educ

## −0.209650 - 10 additional years of education decreases the number of children by ~ 2


# f) Suppose that, given the level of education, your aim is to determine
# how age affects the fertility rate. 
# Make necessary adjustments to the model and estimate it. 
# Test hypothesis whether age has significant effect on fertility. 

## model: children = constant + educ + age

ols children 0 educ age
freq age --plot=display
series uhat1 = $uhat 
freq uhat2 --plot=display # plot residuals

# g) Test hypothesis whether age and education are jointly statistically significant. 
# Clearly state the null and alternative hypothesis and the result of the test.   
H0: B_educ = b_age = 0
H1: B_educ <>0, B_age <> 0