3
$\begingroup$

I have a question

How would you interpret the odds ratios for a continuous variable?

$\endgroup$
  • 2
    $\begingroup$ It might help if you clarified the context. It seems like you're talking about a continuous predictor in logistic regression (as I've assumed in the answer below) but more information would be helpful. I have the same comment regarding your interaction question - stats.stackexchange.com/questions/22036/… $\endgroup$ – Macro Feb 1 '12 at 2:27
9
$\begingroup$

It's not completely clear from your question, but I'm assuming you're talking about the situation where you have a single binary response $Y$ and a continuous predictor $X$ and fit a logistic regression model:

$$ \log \left( \frac{ P(Y_{i}=1|X_{i}) }{P(Y_{i}=0|X_{i})} \right) = \beta_{0} + \beta_{1} X_{i} $$

Then the odds ratio, $e^{\beta_{1}}$, is the odds ratio associated with a one unit increase in $X$.

Essentially you can think of it as the odds ratio between $Y$ and the dummy variable $B$ defined such that $B=1$ if $X=x+1$ and $B=0$ if $X=x$.

Note: The assumption underlying the logistic model is that this odds ratio does not depend on $x$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.