itemparm: extract lower-order structural (item) parameters

Description

Function to extract various item parameters, including "itemprob" for item success probabilities of each reduced attribute pattern, "catprob" for category success probabilities of each reduced attribute pattern, "LCprob" for item success probabilities of each attribute pattern,"gs" for guessing and slip parameters, "delta" for delta parameters, "rrum" for RRUM parameters when items are estimated using RRUM. Standard errors can be estimated if withSE = TRUE. See GDINA for examples.

Usage

itemparm(object, what = c("catprob", "gs", "delta", "rrum", "itemprob",
  "LCprob"), withSE = FALSE, SE.type = 2, digits = 4, ...)

Arguments

object

estimated GDINA object returned from GDINA

what

what to show; It can be "itemprob" for item success probabilities of each reduced attribute pattern, "catprob" for category success probabilities of each reduced attribute pattern, "LCprob" for item success probabilities of each attribute pattern, "gs" for guessing and slip parameters, "delta" for delta parameters, "rrum" for RRUM parameters when items are estimated using RRUM. The default is "catprob".

withSE

show standard errors or not?

SE.type

Type of standard errors. Can be 1, 2 or 3, indicating outer product of gradient (OPG) estimates based on itemwise, incomplete or complete information matrix. See Philipp, Strobl, de la Torre, & Zeileis (2016). Currently, the OPG method based on the complete information matrix assumes that all latent classes are identifiable.

digits

how many decimal places for the ouput?

...

additional arguments

References

Philipp, M., Strobl, C., de la Torre, J., & Zeileis, A.(2016). On the estima-tion of standard errors in cognitive diagnosis models (Working Papers). Fac-ulty of Economics and Statistics, University of Innsbruck.Retrieved from http://EconPapers.repec.org/RePEc:inn:wpaper:2016-25