I am examining BMI effect on wage. I have included a lot of explanatory variables in the regression and all (most) coefficients were significant, but once I add ability which correlates with education (already in regression) slightly (34%) BMI measures turn insignificant.
Here is the question, what does it mean if BMI turns insignificant once we add another variable in the regression?
Estimated by OLS, cross sectional. Adj R 0.60, F=160.
PS: I added a variable and variable of interest turned insignificant, not vice versa. Thank you for all the examples below, they address the same topic, but different issue.