0
$\begingroup$

ordered logit model

I am trying to analyse my experimental survey results, where people are given one of two frames at random. 1.Gain is one of the frames. The dependent variable is the difference between whether people thought it was a good idea after the frame and before the frame.

I need help on how to read the coefficient

$\endgroup$
  • 1
    $\begingroup$ It would be better if you can post the summary as a code snippet, rather than an image. Would be more clear. $\endgroup$ – Dawny33 Aug 20 '15 at 10:54
  • $\begingroup$ Although written using R code, you may want to look at this answer: Interpretation of ordinal logistic regression. $\endgroup$ – gung Aug 21 '15 at 18:37
1
$\begingroup$

The coefficients are not really interpretable since they give you the marginal effect on the latent scale, where the true metric is unknown since they are identified up to scale only.

You probably want to use the margins command after you fit the ologit, like this: margins, dydx(Gain).

This will give you the additive marginal effect on the probability for each of the three outcomes from seeing the gain frame.

You can also calculate $\exp(\beta),$ which gives you the multiplicative marginal effect on the odds of higher outcome compared with the lower outcome. Personally, I find these harder to wrap my brain around. You can get this by adding the or option at the end of the ologit. In this case, $\exp(.4554101)=1.5768199$, which means that the odds of being in the top category versus the combined middle and low are about $1.6$ times higher for those who saw the gain frame compared to the other one. You can also say the same thing about top and middle compared to the bottom category.

$\endgroup$

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.