Regression with Categorical Predictors Flashcards
Categorical Predictor
A predictor measured on a categorical (or nominal scale)
These variables classify observations into groups and have a limited number of levels.
Ordinal Data
Placed in order e.g. sport performance
Interval and ratio data
E.g. temperature
Ratio has a true 0 point
Dummy Coding
Used when you want to assess whether a categorical variable predicts an outcome.
Categories are represented by the means of the values 0 and 1.
What is meant by a baseline category in dummy coding?
A category against which the other categories should be compared.
Assign the value of 0 to the baseline and the value of 1 to the category which will be compared against baseline.
What is the B value of a categorical predictor with two categories equal to?
The difference between the mean scores obtained on the outcome by the two categorical groups
When there is a categorical predictor with two categories, what does the intercept in the regression equation correspond to?
The mean of the baseline category (category to which we have assigned the value of 0)
What role does a t-test play when a predictor has two categories
Tests the null hypothesis that the regression line is flat (b=0) i.e. changes in the value of the predictor do not lead to changes in the value of the outcome.
If t is significant then the baseline category and the other category are significantly different.
Only look at when have 2 categories.
Explain the procedure when a predictor has more than two categories
- Numb of dummy variables: Number of categories - 1
- Choose one category as baseline and assign 0 to that
- First dummy variable: assign value of 1 to first category that you want to compare against the baseline and assign 0 to all others.
- Second dummy variable: Assign value of 1 to the second category that you want to compare against the baseline category and assign 0 to others.
Which SPSS outputs are relevant for regression with categorical predictors?
Model summary
ANOVA
Coefficients table
