USING MULTI-FACET RASCH MODEL (MFRM) IN RATER-MEDIATED ASSESSMENT

Abstract

This paper is an introduction to the MFRM. It is intentionally meant to be simple with an attempt to avoid the sophisticated mathematical equations on which the calibration of the involved facets is based, wherever possible, so that little mathematical background does not obstruct the understanding. This paper aims at introducing the lay reader, who is involved in language performance assessment with no expertise, yet striving for objective assessment, to the Multi-Facet Rasch Model (MRFM) approach (Linacre, 1989). The reader will learn about the MFRM, its conceptual foundations and development, its powerful features, its implementation in rater-mediated assessment contexts, and the interpretation of its main statistical indices pertaining to the facets under investigation. The data used to illustrate FACETS analysis throughout the paper are part of a larger data set collected in the study (Medvedev & Bahrouni, 2013) funded by Sultan Qaboos University.