I am building a logistic regression model and have one polytomous independent variable. Having converted this variable into a series of dummy variables (with a baseline group of course), I now find that some of the dummies are statistically significant and some are not. Is it appropriate to simply remove the statistically insignicant variables from the model? If not, what are the issues with doing so?
Thanks much! . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
