Comparing variable order
The order of predictor variables can be important, especially if predictors are correlated. This is because changing the order of correlated predictor variables can change the estimates for the regression coefficients. The name for this problem is Multicollinearity.
During this exercise, you will build two different multiple regressions with the bus data in order to compare the importance of model inputs order.
First, examine the correlation between CommuteDays
and MilesOneWay
.
Second, build two logistic regressions using the bus
data frame where Bus
is predicted by CommuteDays
and MilesOneWay
in separate orders.
After you build the two models, look at each model's summary()
to see the outputs.
This exercise is part of the course
Generalized Linear Models in R
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Run a correlation
___(bus$___, bus$___)