Rasch goes open Source: The R package eRm for the computation of extended Rasch models

  • Patrick Mair (Ko-Autor*in)
  • Marco Maier (Ko-Autor*in)

Aktivität: VortragWissenschaftlicher Vortrag (Science-to-Science)

Beschreibung

R is an Open Source environment for statistical computing and is considered as the "lingua franca" of computational statistics. Over the last few years, several packages from the area of Psychometrics in general and IRT in particular have been developed. One of them is the eRm package (Mair & Hatzinger, 2007) for computing extended Rasch models. This package implements the Rasch model, the linear logistic test model (LLTM), the linear logistic model with relaxed assumptions (LLRA), the (linear) rating scale model (RSM and LRSM), and the (linear) partial credit model (PCM and LPCM). Conditional maximum likelihood estimation (CML) is used for item parameter estimation. Various model diagnostics/tests such as Andersen's LR-test, Ponocny's nonparametric tests, item-fit and person-fit statistics, graphical model tests, etc. are available.

In this talk we emphasize the flexibility of R and eRm, respectively, for performing Rasch analysis. It allows to create highly customizable high-quality plots, is very powerful for simulations studies and allows a post-processing of the results, also in connection with other packages. eRm is available for free at http://CRAN.R-project.org/package=eRm.
Zeitraum13 Juni 201016 Juni 2010
EreignistitelRasch Conference
VeranstaltungstypKeine Angaben
BekanntheitsgradInternational