Scenario: Crammer Nation University wants to develop a regression equation to predict the "Number of Recruits" a given fraternity will receive this rush season given the "Parties Thrown" by the fraternity last year and the average "GPA" of that fraternity. They take a random sample of 52 fraternities on campus, resulting in the regression output below.
Imagine you want to add "Chapter Meeting Location" into the model, which has the following values: chapter-house, annex-house, classroom. How many dummy variables would you need?
bAnnex is the change in predicted response ("Number of Recruits") if a fraternity hosts chapter at an annex-house. Annex equals 0 (if the chapter doesn't host at an annex-house) or 1 (if the chapter does host at an annex-house). bClassroom is the change in predicted response ("Number of Recruits") if a fraternity hosts chapter in a classroom. Classroom equals 0 (if the chapter doesn't host in a classroom) or 1 (if the chapter does host in a classroom).
Each dummy variable is comparing itself to the baseline.
Remember when we said above...
A fraternity who hosts chapter at an annex-house gets bAnnex / Classroom more / less "Number of Recruits" than one who hosts at their chapter-house (baseline).
Imagine that you’re building a model to predict a student’s exam score based on the number of hours that they studied. You want to add the location that they studied as a variable into the model, with 3 eligible values: dorm, library, coffee shop.
How many dummy variables need to be added into the regression to represent the student’s study location?
Imagine that you’re building a model to predict a student’s exam score based on the number of hours that they studied. You want to add the location that they studied as a variable into the model, with 3 eligible values: dorm, library, coffee shop.
With “dorm” as the baseline for your study location dummy variables, you build the following multi-linear regression equation:
