fscores and fixing parameters

[Aaron Kaat]

I want to use fscores to get EAP estimates for a scale that already has calibrated parameters published. fscores() requires a object of mirt-class, so I thought I would run mirt() with the parameters fixed to their known values, free the latent mean/covariance, and expect rapid convergence. That doesn't work. I get this error:

Error in UpdatePrepList(PrepList, pars, random = mixed.design$random,  : 

  pars input does not contain the appropriate classes, which should match pars = 'values'

Please help me understand my error. I reproduce the code below. Why doesn't the pars file match?

Thanks for your help!

Aaron

ASR_mirt <- c(t(ASR)) #convert matrix to vector by rows

scorable <- c("asr1","asr4","asr5","asr7","asr9","asr11","asr13","asr14") #set items to score

parnums <- mirt(dat[,scorable],model=1,itemtype="graded",quadpts=81,pars='values')

parnums$value <- c(ASR_mirt,0,1) #known values, starting mean, starting cov

parnums$est <- c(rep("FALSE",length.out=24),TRUE,TRUE) #fix known values, free mean/cov

parnums #confirm formatting

RADAR_mirt <- mirt(dat[,scorable],model=1,itemtype="graded",quadpts=81,pars=parnums)

[Phil Chalmers]

Hi Aaron,

See this CV post for an example of how to do this, even with the newer wrapper functions in mirtCAT: http://stats.stackexchange.com/questions/34119/estimating-ability-using-irt-when-the-model-parameters-are-known/48105#48105

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.
For more options, visit https://groups.google.com/d/optout.