*Pardon as I've asked this question in another thread but I am not getting resposnes, so posting here*
If I have tested the relationship between my study variable and the outcome variable with chi-square and not-significant. Then I went ahead and ran my full model with control variables and explanatory variables and one of my study variables turned up significant in the outcome. Would you consider this valid, or invalid? The study variable is different than the one testing in chi-square.
Thank you,