manuelreif / PP

Person Parameter estimation (1,2,3,4PL model) and GPCM

PP

Version 0.6.3 is on cran!

This R package provides Person Parameter estimation for the 1,2,3,4PL model and the generalized partial credit model. This package will soon be uploaded to cran!

To install this package from github, install devtools first.

devtools::install_github("jansteinfeld/PP", build_vignettes = TRUE)

What's NEW?

• fixed knitr issue 1864 by declared 'rmarkdown' package as a dependency
• Jan Steinfeld is now the maintainer of the package, the main feautres and whole idea of the package is from Manuel Reif
• Jan Steinfeld is now on board!
• Jan programmed the great Pfit() function, to estimate a bunch of fit indices!
• Use the brandnew PPass() function, to estimate Person Parameters and fit indices in one step!
• We added a real-world dataset, which contains responses to a adaptive intelligence test. Load the data and start trying the functions of PP.
• You can now fit your model with the great eRm package, and put the object into the PPass() function to estimate person parameters and person fit indices.

Example

Here is a small example of the new PPass() function.

> data(pp_amt)
> 
> d <- pp_amt$daten_amt
> 
> rd_res <- PPass(respdf = d, 
+                 items = 8:ncol(d),
+                 mod="1PL",
+                 thres = pp_amt$betas[,2], 
+                 fitindices = "lz")
Estimating:  1pl model ... 
type = wle 
Estimation finished!
> 
> head(rd_res)
   estimate        SE        lz   lz_unst
1 1.4098582 0.4177663 -0.631025 -0.613445
2 0.3515598 0.3728682  1.229631 -0.627897
3 1.0745005 0.3741312  1.097723 -0.627596
4 0.3204690 0.6900113  0.981567 -0.494311
5 1.3879169 0.4643963 -0.203400 -0.600039
6 0.9457360 0.4184997  0.573956 -0.615722