If a Researchers in an Observational Study Attempts to Control

Multiple Choice

If a researchers in an observational study attempts to control for a confounding variable, is it more likely that the resulting confidence interval will be wider or narrower than the confidence interval when he/she didn't control for the confounding variable?

A) It will probably be wider.
B) It will probably be narrower.
C) It will have to be disregarded; you can't adjust or control for confounding variables after the fact.
D) There is no way of knowing.