Page 78 - BUKU SYNOPSIS

answers indicating overall language proficiency. However, there is a

distinction when evaluating language proficiency in speaking and

writing abilities since learners must showcase their language skills

through performance evaluations involving speaking and writing tasks.

The objectivity of the assessment of learning outcomes is

significantly influenced by the quality of the instrument or the test.

Only assessments based on reliable measurement data can be used to

determine the achievement of teachers and students (Brown and

Abeywickrama (2018); Mardapi (2017)). The test developer can use

either classical response theory or modern theory, also referred to as

CTT (Classical Test Theory) and IRT (Item Respond Theory), to

evaluate the quality of multiple-choice or true-false tests in the context

of language assessments. A score of one indicates that the answer is

right, and a score of zero indicates that the answer is incorrect for the

test item. When speaking and writing ability proficiency is assessed

through a performance test, however, the situation is different. A

grading scale or rubric will be more appropriate for

assessing performance (Engelhard et al., 2018; Finch & French, 2019).

According to Knoch (2009), the rating scale is the central aspect of the

performance assessment writing and represents the "de-facto test

construct (p.14)."

The rubric functions as a pedagogical instrument teachers

employ to maintain accurate and reliable assessments of the quality of

73