Correlating variable added to logistic regression changes significant predictors

#1
Hi, when I added a new correlating variable with my DV to the logistic regression model (which wasn't in itself a significant predictor), it changed one of the previously significant predictors to a non-significant predictor.
Is this a limitation of a regression model? Does anyone have a good link for an article or book that will explain this?
Thanks
 

hlsmith

Less is more. Stay pure. Stay poor.
#3
You could also read about bias/variance trade off. Of note, did it change the estimate that much or did it just make the SE larger? There is a difference there. Also, this wouldn't be considered a limitation of regression, because the model is now adjusting for another variable. It would be a strength of regression.