Regression coefficient and SE = 0

#1
I have a continuous variable in my regression model that is normally distributed. It has a statistically significant relationship with my continuous outcome variable, but the regression coefficient and standard error are both 0. What should I do in this situation? What would cause the coefficient and SE to be 0?
 

hlsmith

Not a robit
#3
Are they really "0" if they have a "statistical significant relationship". Please post output from the regression model and make sure the sample size is apparent.

Thanks
 

hlsmith

Not a robit
#5
can you take a screen shot of the full output and perhaps tell use about data (e.g., sample size)?

Why are you exp your coefficient, you do that with link functions like in logistic regression. You have two continuous variables, so I had imagined you were running a linear or generalized linear model. Are you just using the wrong model here?
 

hlsmith

Not a robit
#7
I will also note that even though the output looks like estimates are null, that doesn't mean that beta may not be 0.00006932 and SE is 0.00001. That would be weird, but theoretically possible. Also what were you expecting for the odds ratio?