Present your results in a table.
Create a figure that includes a scatter plot of your x and y and also shows the predicted probability.
Tell us what you can interpret from the coefficient.
Give the marginal effect of an increase in x on y when x is equal to 25, 50, 75.
Explain why the marginal effect changes, both mathematically and intuitively.
How many predictions did you get correct with this simple model?
Write a program in the Stata ML environment to replicate the results from the first step in the program.
Chapter 9 Question Part 2 (Categorical)
In this section, you will create a dataset and work on modeling how students choose sections for my class in the spring. You will want to create two datasets: 1) a pseudo-individual level dataset (i.e. if there are 15 people taking section 1, you will add 15 rows to the data set with “1” for their section choice) and 2) a data of characteristics of each of the section (i.e. time and day of week). You will need to expand the first dataset and merge it with the second, as we did in class, and then run a CLOGIT to answer the following questions. Below is a table with information on each section:
Section Number Students Enrolled Time of Day Day of Week Modality
1201 3 9am Friday AR
1202 4 10am Friday AH
1203 10 11am Friday AH
1204 5 12pm Friday AH
1205 16 2:30pm Wednesday AR
1206 19 9:00am Thursday AR
1207 15 10:30am Thursday AR
1208 3 7:30pm Thursday AH
Run a CLOGIT on student choice of section with at least three explanatory variables based upon the information in the table above. Present your results in a table and discuss the sign and significance of each coefficient.
Choose one of your explanatory variables, and then present the marginal effect of a change in the value of this variable for Section 1201 on the probability of a student choosing section 1201, as well as the impact of this change on the probability of students choosing all other sections.
Suppose sections 2,4,6 and 8 are cancelled. According to your model, how will students be distributed among the remaining 4 sections?
Chapter 11 Question
For this question you will need to clean some data. Take the ur.dta dataset and merge it with the data from the URL listed above so that you have data on the average unemployment rate in a quarter and quarterly GDP growth. Keep data from 1950 to present that you can match (i.e. you can’t use October 2017). You will want to use the commands “collapse” and “merge” here. With this cleaned data, answer the following questions:
Using only data from 1950 through 2016, test for stationarity in both variables. Ignore drifts and trends.
Suppose you are asked to regress UR on GDP growth. In order for this regression to be meaninful, what needs to be true?
Is UR stationary?
If not, how many lags do you need to include for it to become stationary?
Is GDP growth stationary?
If not, how many lags do you need to include for it to become stationary?
Can you run a meaningful regression of UR and GDP given what you have found above and any other evidence you may want to explore? If not, run a model with transformations of the variables that would be meaningful.
Run a VAR of UR on GDP growth.
Try this with 1-3 lags. Pick the best model according to the BIC and report only those results.
Does GDP growth Granger cause UR?
Does UR Granger cause GDP growth??