Page 178 - BUKU SYNOPSIS

1) Analyzing the qualities of items/construct of the assessment

instruments

a) A writing Test

This test was designed to assess the student's ability to write an

argumentative text. When measuring learners' skills in oral and written

communication, rating scales or rubrics are used. As a result, the rating scales

or scoring rubrics must be aligned with the test grid to ensure the construct

validity of the test. The scoring rubric used in assessing spoken and written

language, according to McNamara (1996), is a visual depiction of the test.

1) Construct Validity

The validation of construct or item fit for evaluating writing skills,

particularly in the context of composing argumentative texts, has been

demonstrated using Many Facet Rasch Measurement (MFRM) modeling with

the aid of the facet application. The concept of item fit pertains to the extent

to which the components within the assessment rubric effectively and

accurately evaluate students' writing abilities, while minimizing any potential

misunderstandings or biases introduced by the assessor. The measurement of

item fit encompasses four criteria, namely infit Mnsq, outfit Mnsq, outfit Zstd,

and point measure correlation.

173