skip to main content skip to footer

Recent Additions to the MGROUP Set of Programs IRT IALS NAEP

Author(s):
von Davier, Matthias
Publication Year:
2005
Report Number:
RM-05-02
Source:
ETS Research Memorandum
Document Type:
Report
Page Count:
16
Subject/Key Words:
Conditioning Model, Item Response Theory (IRT), Latent Regression, Subgroup Statistics, MGROUP, International Adult Literacy Survey (IALS), National Assessment of Educational Progress (NAEP)

Abstract

The MGROUP set of programs is used in large-scale educational assessments such as National Assessment of Educational Progress (NAEP) and International Adult Literacy Study (IALS) to estimate consistently subgroup statistics based on sparse matrix samples of item responses and a complete set of background variables for each respondent. This process is sometimes referred to as conditioning in the literature, and the models applied for this conditioning may be viewed as latent regression models, focusing on the assumed latent population structure or models for marginal estimation of (sub)population characteristics. This paper summarizes modeling and estimation options implemented in YGROUP, a recent addition to the MGROUP set of programs. Where available, the documented additions will be related to simulation studies that demonstrate the comparability of results with results from existing programs or demonstrate the perceived advantages of the implemented changes.

Read More