In regression estimator, if Y is study variable and X is an auxiliary variable. The assumption is that X is always stable throughout the process. Because X is always stable may not be correct in most real cases. Please suggest how in real scenario, X will be stable throughout the process?
Stability of the X variable is important for the equation to make sense therefore it should be tested and a binary test or Chow test can be used.
You can look into the following book " Applied Regression Analysis " by Draper and Smith, in this book there is discussion about regression analysis under the situation when both the regressors and response variables are treated stochastically.
Stability of the X variable is important for the equation to make sense therefore it should be tested and a binary test or Chow test can be used.
- Badges
- Science topic