In the near future I will release MiscPsycho, a package that contains
various functions useful for applied psychometricians. I would like to
include some data sets for distribution in the package, but have not
created any of these on my own, but have used data distributed in other
packages such as the LSAT data in the ltm package.
Is it appropriate for me to distribute a data set in the package I
develop even though it originally came from another package? Or, is it
more appropriate for my package to depend on the package where the data
originally live?
The functions planned for the current distribution of MiscPsycho will
include:
1) JML estimation of Rasch and Master's partial credit model
2) Various IRT statistics (e.g., item information and test information
functions)
3) Various "Classical" statistics for item analysis
4) Coefficient alpha
5) Score conversion tables
6) Classification consistency analysis
7) Procedures for linking scales using mean equating
Future development will include
* 2PL, 3PL, and GPCM also using JML
* Stocking-Lord procedure
* Various methods for assessing DIF, including MH
* Bayesian scoring methods (e.g., EAP)
The JML procedure is not the world's best technology as MML is commonly
preferred. However, that is already implemented in the ltm package and
these procedures are provided only for those who want to JML option.
Thanks,
Harold
[[alternative HTML version deleted]]
______________________________________________
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
