How to interpret LD standardised residuals

[Nikolaos Tsigilis]

Hi Phil,

In examining the local independence assumption recent updates of mirt  provide standardisation of the residuals.  I am wondering how to interpret these residuals.  If I am not wrong one way is to search for values above 1.96 or 2.57. However, in his article Toland (2014) using the IRTPRO suggests that residuals with absolute values above |10| denote violation of the assumption. What is your opinion?

Best regards,

Nikos

[Phil Chalmers]

Cutoffs are cutoffs; I don't enjoy them but I understand that some researchers like to rely on arbitrary lines in the sand for simplification. You could make similar claims for the Cramer's V outputs from mirt, though a simulation study would have to be run to determine reasonable cutoff locations for reporting. If it helps though, you can use the residuals(..., approx.z) argument to obtain the same standardized residuals in which the |10| cutoff would apply.

Phil

--
You received this message because you are subscribed to the Google Groups "mirt-package" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mirt-package...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mirt-package/8b4b9981-6be7-46cf-8a5b-09e4c4bd0023n%40googlegroups.com.