turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- Incidence Risk Ratio (IRR)

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

06-27-2017 12:33 PM

Hi everyone

I am trying to get the incience risk ratio (irr) for each of my independent variables while also using the zero inflated negative binomial regression. How can i do this as the output for ZINB does not give the irr.

Thanks

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

Posted in reply to AY_MANSA

06-29-2017 05:00 AM

In my opinion, you should forget this idea and fit data with Poisson regression.

This is because the likelihood obtained from observing person time and events is the same function of the parameters as when you have poisson distributed counts. Therefore, Poisson regression is just used as a trick to maximize the likehood function and thereby find the right estimates of the IRR. It doesnt matter that the fitstatistics shows a misfit, as the data was not Poisson distributed anyway.

Negative binomial regression will only result in biased standard errors.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

Posted in reply to AY_MANSA

06-30-2017 02:03 PM

A negative binomial or ZINB model is for a count response. A risk ratio is a ratio of event probabilities, so that assumes that the response is binary, not a count. Assuming your response is really a count, then you might want to estimate a RATE ratio. This note discusses estimating rates and rate ratios with a count model. This note talks about estimating rates using zero-inflated count models.