Kucharssim
commented
2 years ago Dear @khh200,
thank you for your request. Adding robust standard errors should be feasible, we will soon have HC0-HC5 methods in ANOVAs under order restrictions. Adding this to regression would be a natural future step. I don't know what is the roadmap for order restrictions with linear regression though, @EJWagenmakers? Of course, adding the robust standard errors can be implemented separately from the order restrictions, but implementing them at the same time makes it somewhat more straightforward.
I am not really familiar with the Hausman test, but since you write that's for the random effects models, can @FBartos perhaps look into this request?
Description
JASP should include HC0:HC3 robust standard errors, cluster robust standard errors, and the hausman test under the random effects models.
Purpose
To allow for a wider array of researchers to engage with JASP. In econometrics, random effects models are almost never used due to violations of endogeneity.
Use-case
It provides psychologists alternative means of dealing with clustered data and attracts a new field for JASP usage.
Is your feature request related to a problem?
No.
Describe the solution you would like
I would like JASP to incorporate the various robust standard errors for regression from a package like sandwich and lmertest. Additionally, I would like the hausman test added so that the endogeneity assumption of HLM may be assessed (from the plm package for example).
Describe alternatives that you have considered
No response
Additional context
No response