Hi all! I would like to fit a simple regression model to a number of observations. However, I have reasons to think that maybe the relationship I am testing is significant only for a number of them. Is there some automated method to test the model for all possible data combinations? Thank you! Dimitris Moutopoulos