Assignment 1

Name                                                                           Score

Using the data (Nigeria3.dta) for all the analyses. Create a do file and save all your outputs, each with a file name started with YOUR LAST NAME_. Submit your Do file and Outputs file along with your answers to the following questions.

In this assignment, we are interested in the determinants of fertility willingness.

Task 1. To make a descriptive table for all the variables you may use in the analysis, including your dependent variable fertility willingness measured by “ideal number of children (V613)”, as well as potential predictors V024, v106, v160 v190 v212 V404 v701 b4, etc. (Hint: To report mean and standard deviation for continuous variables, and frequencies for categorical variables, remember to include missingness information) (10 points)

Task 2. To present your data graphically (10 points in total),

1. First to create a box plot for V613 (5 points)
2. Then to create box plots that show the distribution of ideal number of children, based on the categorical variable region. What does the graph (b) tell you about the relationship between fertility willingness and region? (5 points).

Task 3. To estimate the education effect (20 points in total)

1. First, estimate the education effect of wife. Please interpret your findings, including the coefficient and model fitness (5 points)
2. Then include both the effects of the wife and the husband. What are your findings? (Has the model fitness changed? Have you noticed any change to the coefficient for the wife’s education, and what might be the reason for any change seen?) 5 points)
3. What is the correlation between the wife’s education and the husband’s education?
4. If the two categorical variables are treated as though they were continuous variables, what does the correlation tell? (Hint: c.) (5 points)
5. Then correctly treat the two categorical variables as categorial variables (Hint: i.). What analysis do you use to test the relationship between the two categorical variables? What do the results tell?  Compare the results to the estimated correlation when you treated them as continuous variables. (5 points)

Task 4. To study the effect of religion (15 points in total)

1. Does Muslim families have higher fertility compared to Catholic families? If so, by how much? (5 points)
2. Do Catholic families have higher compared to other Christian families? How can you tell? (5 points)
3. Now, include education in the model. Does education mediate the religion effect on fertility, and in which way? (5 points)

Task 5. To test age of respondent at first birth and fertility willingness (25 points in total)

1. What is the correlation between age and fertility? (5 points)
2. Make a scatter plot of two variables. Pay attention to put ideal number on Y axis! (5 points)
3. Run a simple linear regression of fertility willingness on age of first birth. (5 points)
4. Test whether there is nonlinear age effect (Hint, include an age square term in addition to age). What is your finding? Is the age effect nonlinear? (5 points)
5. Estimate a standardized coefficient for the age effect. What does the coefficient tell? (5 points)

Task 6. To examine other potential variables (10 points in total)

1. Does the sex of the child matters? What is your hypothesized effect? Do you think the results make sense? (5 points)
2. How about final say on making large decision? (Hint: In your analysis, make the category with largest number of cases as your omitted category) (5 points)

Task 7. To build a multiple linear regression.

1. What variables do you want to include? Why?
2. Build your multiple linear regression. Interpret your findings and comment on the model fitness. (10 points)