Variable changes from not significant to significant, don't know why, please help

Question

When doing a regression on 3 variables to see how they effect electricity usage one variable comes out as not significant but with the correct expected relationship. When I put these 3 variables and add 6 others to it and run the regression again the variable now becomes significant (still with correct relationship), am I doing something wrong or should I change something? Thanks

Answer 1

The only thing you are doing wrong is placing too much emphasis on "significance".

There is no reason that what you are describing can't happen. The relationship between an IV and a DV changes when you add other IVs; this is part of what controlling for other variables does.

Without knowing what your variables are, what your study is about, what your sample size is, what the effect sizes are in the two models, it's hard to say anything else.