Answer-first summary for fast verification
Answer: You should perform t-tests on each X variable to establish whether there’s any of them whose effect on Y is not statistically significant
The F-test in multiple regression establishes whether at least one of the independent variables is statistically significant (i.e., whether the overall model explains a significant amount of variation in Y). However, it does not specify which individual variables are significant. After obtaining a significant F-test result, the next logical step is to perform t-tests on each individual X variable to determine which specific variables contribute significantly to the model and which ones may be insignificant and could potentially be removed.
Author: Nikitesh Somanthe
Ultimate access to all questions.
Suppose you performed the F-test on a multiple regression model and established that a significant amount of variation in the Y variable is explained by the set of X variables; then:
A
You should make a transformation of the Y variable
B
You could perform another test with an indicator variable so as to establish the significance level of the test
C
You should discard the initial model in favor of a different one with more X variables
D
You should perform t-tests on each X variable to establish whether there’s any of them whose effect on Y is not statistically significant
No comments yet.