Does anyone have any advice about how to set an appropriate ROPE range for the gamma portion of a hurdle-gamma model? Is the usual ± .1SD*Y appropriate (I know it’s not for the hurdle part of the model, but am not sure for the gamma part)? Thanks in advance for any advice. Cheers, Emily

Welcome to the Stan forum!

I am not sure if your question is more technical or conceptual.

Conceptually, I think a ROPE (region of practical equivalence) should be defined based on domain knowledge. To get an idea about what the hurdle part of a hurdle-gamma plays, it would be useful to know what the data generating process is, and what the research question is.

thanks for welcome and the response. I’m not sure if the question is conceptual or technical - a bit of both I guess. I understand how to set a ROPE based on domain knowledge for a general linear model with a Gaussian link function. I also know how to set it for a logistic model with a logit link. What I don’t know is how to do it for a gamma model with an inverse link function. My mathematical capabilities break down at this point.