ABSTRACT
Different observers may record different observations for the same event or object. Testing services and examination boards often employ multiple raters to code each response. Most researchers were introduced to the Many-facets Rasch model (MFRM) through Linacre’s 1989 book Many-facet Rasch Measurement. The output of the MFRM gives information regarding the MFRM estimates of the raters, the questions and the coding categories of the rating scale. The dataset consists of only multiple columns. The first column is pid and contains the identifier of the respondents. The second column is rater and contains the identifier for the raters. The only change in the interpretation of the model output is that—in addition to deltas, taus and thetas—the model estimates rater-specific alpha values.
