First off LTM is an amazing package and I'm really grateful to be able to use it so thank you!
Now, my issue, I'm trying out different configurations with gpcm, in an attempt to cross validate difficulty and discrimination estimates coming from the PASRCALE software platform (the default settings produced difficulty estimates that were an average of 0.24 off).
When I've tried to change the the optimizer to 'optim' I get the following error:
Error in if (any(ind <- out == 1)) out[ind] <- 1 - eps :
missing value where TRUE/FALSE needed
A quick search revealed that this issue has come about before, but the solution of removing items with 2 or less responses does not seem to be relevant here because I can subset my data to include full responses to each item and I still receive this error.
If there's any help or information you can provide to resolve this issue I'd be very grateful,
Thanks
Hi Dr. Rizopoulos,
First off LTM is an amazing package and I'm really grateful to be able to use it so thank you!
Now, my issue, I'm trying out different configurations with gpcm, in an attempt to cross validate difficulty and discrimination estimates coming from the PASRCALE software platform (the default settings produced difficulty estimates that were an average of 0.24 off).
When I've tried to change the the optimizer to 'optim' I get the following error:
A quick search revealed that this issue has come about before, but the solution of removing items with 2 or less responses does not seem to be relevant here because I can subset my data to include full responses to each item and I still receive this error.
If there's any help or information you can provide to resolve this issue I'd be very grateful, Thanks