[R-sig-ME] Selecting a family in glmmTMB for continuous proportion data that includes 0 and 1

[April Martinig]

Hello,

I’ve noticed several questions about fitting proportion data with glmmTMB.

Every question I’ve encountered has been focused on 0 < x < 1. However, my data included 0 and 1. I am currently using "family=beta_family(link="logit"), ziformula= ~1” after making the 0s slightly non-zero (0.00001) and the 1s slightly smaller (0.99999) - I tried the solutions under zero inflation continuous data https://bbolker.github.io/mixedmodels-misc/glmmFAQ.html#continuous-data.

However, is there a way to appropriately fit the data as is (true zeros and true ones)?

I’ve attached an image of the distribution of the data.

Any suggestions would be much appreciated!

Take care,
April

—
Dr. April Martinig, Ph.D., M.Sc.
Postdoctoral fellow
University of New South Wales
martinig.weebly.com<http://martinig.weebly.com>
“I do not accept the conventional order of things as a given."


-------------- next part --------------
A non-text attachment was scrubbed...
Name: plot.pdf
Type: application/pdf
Size: 96651 bytes
Desc: plot.pdf
URL: <https://stat.ethz.ch/pipermail/r-sig-mixed-models/attachments/20230604/35bfa505/attachment-0001.pdf>