Exercise

Section 1 (50 marks):

Instruction: Download the World Bank Development Report Data on Blackboard and run the following regressions. Please interpret your results carefully and independently.

Questions:

i.Regress the illiteracy rate on per capita GNP in 1997. Report the coefficient on per capita GNP and its standard error; what do they tell you? Is the sign of the coefficient what you expected? Explain briefly. What is the t-statistic for this coefficient, and what does it tell you? Interpret at the 95% confidence interval. (15 marks)

ii.Regress the infant mortality rate in 1997 on GNP per capita in 1997. Is the coefficient on per capita GNP significantly different from zero? How do you know? Interpret the coefficient in terms of a $1000 difference in per capita GNP. (15 marks)

iii.Regress the infant mortality rate in 1997 on the illiteracy rate. Graph a scatter plot of the data as well as the regression line. (Use the “graphics” option from the Stata toolbar. Stata will output the code as well as the graph. You should then try typing the code in directly instead of the editor and may be save it in your do file. It is much faster) (10 marks)

1

iv.Using the results from part i-iii, what can we say about the causal relationship between illiteracy, infant mortality, and income (GNP)? In other words, tell a story of how illiteracy, income, and infant mortality might affect one another. Be concise but clear about what causes what and why. (10 marks)

Section 2 (50 marks):

Imagine that we have survey data of people’s food expenditure and their basic income and demographic characteristics (e.g. age, education, etc.)

a.Write the linear regression that you would run to answer the question whether people who eat more nutritiously earn more. (10 marks)

b.Can the regression estimate tell us whether eating more nutritiously causes higher income? Why or why not? (10 marks)

c.Do you predict the coefficient of your regression above to be positive or negative?

(10 marks)

d.Do you predict the coefficients b and d to be positive or negative if you regress income or nutrition on education as below? (10 marks)

Incomei=a + b(educationi) + e

Nutritioni=c + d(educationi) + m

e.How might the results from part (c) and (d) affect the interpretation of the regression you ran in (a)? (10 marks)