Assignment 3 Instructions

The purpose of this assignment is to illustrate the appropriate use and interpretation of OLS methodology with different types of explanatory variables. For this assignment, we would like you to identify three different variables that you think predict an outcome, where the predictor variables are of three different types: interval/ratio (continuous), dichotomous (binary), and polytomous nominal (categorical, with more than two categories). Preferably, your outcome variable will be continuous. For each regression you run, please write out the full regression equation in your paper.

First describe your theory as it applies to each predictor variable. Then test the theory for each predictor variable by running the following analyses:

1 Estimate a bivariate regression of your outcome variable on the continuous predictor. Interpret the coefficients (especially the slope coefficient) and other relevant statistics. Include a scatterplot of the two variables with the regression line shown.

2 Estimate a bivariate regression of your outcome variable on the binary predictor, and interpret the key results. Next, dichotomize your outcome variable (explain your approach) and estimate another bivariate regression using the new outcome variable. Compare this with the results from a crosstab of the two dichotomous variables.

3 Compute the means of your outcome variable for each category of the polytomous predictor variable. Then replicate this analysis using dummy variable regression: convert the \(n\) -category predictor variable into \(n-1\) dummy variables and regress your outcome variable on this set of dummies. Interpret regression results and compare them to the output from the mean calculations.

In general, the instructions for papers 2 through 5 apply (of course) to this assignment.

Price: $20.43

Solution: The downloadable solution consists of 11 pages, 943 words and 9 charts.
Deliverable: Word Document