www.phidot.org

Is there a function in RMark that will model average beta estimates? The model.average functions all average real parameter estimates. Thank you.

kellyweintraub wrote:Is there a function in RMark that will model average beta estimates? The model.average functions all average real parameter estimates. Thank you.

Kelly,
I think the answer is no as MARK does not model.average beta estimates either (unless I missed something, which is possible). But, there have been several posts on phidot about this issue and how to do it if I remember correctly.

Bret

Hi, I am trying to do the same thing, but cannot find the 'several posts' you are mentioning here.

I can imagine how to do it manually, but that becomes quite a cumbersome process...

If you are willing to fit all subsets of a global model, see
[url]
viewtopic.php?f=21&t=2478#p7809
[/url]
--jeff

RudyJonker wrote:Hi, I am trying to do the same thing, but cannot find the 'several posts' you are mentioning here.

I can imagine how to do it manually, but that becomes quite a cumbersome process...

Here is a thread that you might find useful on one part of model averaging beta estimates. Based on everything I know about it (which would fit in a beer can) relative to experiences of others on the list (keg size), the how and whether to do it or not is pretty much a gray area...

viewtopic.php?t=996&postdays=0&postorder=asc&start=0