Homework 3: Heteroskedasticity in Regression

Introduction

Some workers participated in a job training program in 1977 and some did not. To estimate the effect of the program on subsequent earnings, consider a regression model that predicts earnings in 1978 (re78) (in thousands of dollars) based on the following explanatory variables:

• re75: Real earnings in 1975 in thousands of dollars (base year 1982)

• re78: Real earnings in 1978 in thousands of dollars (base year 1982)

• educ: Years of education

• age: Worker’s age

• train: Dummy variable equal to 1 if person participated in the training program.

The dataset should be stored in your Rstudio.cloud project files. You can open the dataset with the following call:

load("jtrain3.RData")

Answer the questions below. Type up your answers and submit to the appropriate Canvas assignment folder. Include in your answers (1) the code you used, (2) the output from the code, and (3) your written description of the interpretation of the output as appropriate to answer the question.

Problems

1. Using ordinary least squares, is there evidence that the training program positively influenced earnings?

2. Plot the residuals (vertical axis) against the predicted values (horizontal axis). Is there evidence of heteroskedasticity? If so, describe the behavior of the variance of the residuals.

3. Is there statistical evidence for heteroskedasticity? Conduct the appropriate hypothesis test.

4. Using heteroskedastic consistent standard errors, is there evidence that the training program positive influenced earnings?

Submission

Upload your submission to the Canvas assignment folder titled, “Homework 3 - Heteroskedasticity,” by Monday, April 1, 5:30 PM.

Include in your upload both the .html document and the .Rmd document.