DUMMY
VARIABLE
REGRESSION
MODEL
DISMAS ALEX (MR.)
Dummy Variables
1.1 Describing Qualitative Information
1.2 A Single Dummy Independent Variable
1.3 Using Dummy Variables for Multiple
Categories
1.4 Interactions Involving Dummy Variables
Introduction
Thus far, we have only considered variables
with a QUANTITATIVE MEANING
-ie: dollars, population, utility, etc.
In this lecture we will cover variables with a
QUALITATIVE meaning
-ie: gender, location, race, specific
knowledge or attribute
1.1 Describing Qualitative Information
Any study where an observation has a quality that
can be described as either has/does not have, is/is
not, does/does not etc. can be expressed as a DUMMY
VARIABLE (DV) or BINARY VARIABLE
Ie: -has or does not have a high school diploma
-is or is not male
-does or does not smoke
A way to incorporate qualitative information into
regression model is to use Dummy variables
They may appear as the dependent or as independent
variables
What is Dummy Variable?
Variables that are essentially qualitative in nature or variables that are not
readily quantifiable
Examples: gender, marital status, race, colour, location, nationality etc.
Other names for DV are binary variables, categorical variables, indicator
variables, Dichotomous variables
How to quantify qualitative aspect?
By constructing artificial variables that take on values
of 1 or 0 (ZERO).
1 indicates presence of that attribute
0 indicate absence of that attribute
Example:
Gender =1 if the respondent is female, and 0 if the
respondent is male
Time = 1 if war time, and 0 if peace time
Here variables with values 1 and 0 are called
Dummy variables
WHY Dummy variable in regression?
To include qualitative variables in regression which would otherwise be
difficult to include due to their non-numeric nature.
To enable us to use a single regression equation to represent multiple
groups
Example: If we want to see whether gender discrimination has any
influence on earnings, apart from other factors
A Single Dummy independent variable
Suppose we want to measure the impact of GENDER on WAGES/
Employee compensation
i.e, we are interested to know whether female employees are
discriminated against their male counterparts
Gender is not strictly quantifiable
A Single Dummy independent variable Cont…
Here, we describe gender using Dummy variable
D = 1 if male respondent, and 0 if otherwise.
Interpretation of the model
Interpretation of the model
Cont…
Hypothesis testing
Dummy Variable Trap
-When two Dummy Variables relating to the
same aspect are included, such as male and
female, we cause perfect collinearity
because male+female=1
-this is the DUMMY VARIABLE TRAP that
arises when too many DV’s are included
-The DV Trap can also occur when there are
too many DV’s relative to the different
number of observations
Example of a DV’s Trap
Example
Example
Using dummy variables for multiple categories
Interactions involving dummy
variables
Example
Thank you
