Page 163 - BUKU SYNOPSIS

aims to examine consistency among raters to avoid bias. The researcher

recruited eleven students and two raters for this test.

Item fit using MFRM aims to evaluate the item fit of the analytical

scoring rubric's items with the model, which interprets that the elements

of the writing rubric are appropriate for measuring students' writing

abilities. It can also be used to determine if the two raters comprehend

how to use the developed rubric. Item fit, also shows the validity of the

analytical scoring rubric. MFRM processes data consisting of rater

evaluations of student writing scripts.

Before conducting instrument analysis with MFRM, researcher

needed to ensure the variance explained by Rasch measures, as seen in

Figure 16, where the variance explained by Rasch measures is at 59.99%,

with 0.374, which means the "sufficient" category for further analysis.

The raw variance value explained by measures on the 12 items of the

rating scale greater than the specified criteria, namely 20% (>20%)

(Azizah et al., 2022; Sumintono, 2022). Thus, it can be interpreted that

the items in the rating scale instrument meet the requirements of

unidimensionality or are unidimensional.

Figure 16. Measurable Data Summary

158