代做ECO 480 Midterm代写C/C++程序

Take Home Midterm (ECO 480)

Due on UB Learns Wednesday, 7/31 at 11:59pm

1.) Download and save the data set ‘lead_mortality’. The data description file is included on UB Learns.

a.) Is this data observational or experimental? Why?

b.) Classify this dataset as cross-sectional, time-series, or panel data. Specify the time period(s) and entity(entities) covered by this dataset.

c.) Import the data into excel. (Note: .xlsx file.)

d.) What is the mean infant mortality rate in our sample? What does this average tell us?

e.) What is the standard deviation for infant mortality rate in our sample? Does this standard deviation seem large?

f.) Create a new variable named ‘foreign_percent’ that equals ‘foreign_share’ as a percent. (Note: ‘foreign_share’ is represented as a fraction.)

g.) Rename the variable ‘foreign_share’ as ‘foreign_frac’. Change the label for this variable appropriately.

h.) Create a scatterplot with the typhoid death rate on the x-axis and infant mortality rate on the y-axis. Label the axis and graph appropriately.

FYI typhoid fever results from exposure to Salmonella and was easily spread amongst people in the early 1900s.

2.) Download and save the data set ‘Growth’. The data description file is included on UB Learns.

a.) Import the data into excel. (Note: .xlsx file.)

b.) What is the mean and standard deviation of the variable ‘rgdp60’ in our sample? (Real GDP per capita in 1960 US dollars)

c.) What is the mean and standard deviation of the variable ‘yearsschool’ in our sample?

d.) Create a scatterplot with the average years of schooling on the x-axis and real GDP per capita on the y-axis. Label the axis and graph appropriately.

e.) Run the regression with real GDP per capita as the outcome and average years of schooling as the independent variable. (Note: Standard errors need to be reported robustly.)

f.) Use the format:  to report your regression results. Replace the β’s with your estimates from 2e, and replace the X and Y variable with appropriate names. Report the standard errors for your estimates.

g.) Interpret the  estimate (estimated slope coefficient) from your results.

h.) Which estimates (slope coefficient, intercept, or both) from this regression are significant at the 5% level? How do you know?

i.) Create a variable for the predicted real GDP per capita, based on your estimates

j.) Create a graph with the scatterplot from 2d and the linear regression line you estimated. Please label the axis and graph appropriately.

3.) Download and save the data set ‘birthweight_smoking’. The data description file is included on UB Learns.

a.) Import the data into excel. (Note: .xlsx file.)

b.) Run the baseline regression with birthweight as the outcome and the smoker indicator as the independent variable. (Note: Standard errors need to be reported robustly.) Report these results.

FYI: birthweight is a common outcome when studying infant health because USUALLY unhealthy babies weigh less and healthy babies weigh more.

c.) Interpret the coefficient on the smoker variable from your results in 3b.

d.) Give an example of a factor/influence that will cause omitted variable bias in the regression from 3b. Explain why omitting this variable causes bias (hint: 2 conditions).

(Note: this can be a variable already in the dataset, or a variable not measured in the dataset.)

e.) Run the alternative regression with birthweight as the outcome, and smoker and the number of prenatal visits as the independent variable. (Note: Standard errors need to be reported robustly.) Report these results.

f.) Report the adjusted R-squared for the regression in 3e and interpret this number.

g.) Interpret the coefficient on the smoker variable from your results in 3e.

h.) Suppose we create a variable named noSmoker that = 1 when the woman did not smoke during her pregnancy and noSmoker =0 otherwise. If we include this variable in our regression from 3e, what error would occur? How could we solve this problem?

i.) Report the overall regression F-stat from the regression in 3e. State the null hypothesis that this overall regression F-stat tests. Do we reject or not reject the null hypothesis based on the F-stat and corresponding p-value?