Tahapan Pengembangan Instrumen Evaluasi Maharah Kalam Berdasarkan Kajian Teoretis dan Hasil Penelitian Terdahulu

Evaluating *maharah kalam* (speaking skills) requires instruments capable of measuring speaking proficiency in a valid and reliable manner; however, current practices often rely on assessing comprehension without the use of standardized rubrics. This study aims to outline the stages for developing *maharah kalam* evaluation instruments, drawing upon theoretical frameworks and previous research. Employing a qualitative descriptive approach through a literature review, this study examines theories of language assessment, instrument development models (such as those by Mardapi, the 4D and 3D models, and the ADDIE model), as well as ten studies published between 2021 and 2026. The findings indicate that while existing research typically involves needs analysis, the design of tasks and rubrics, and expert validation, evidence of reliability—particularly inter-rater reliability—is rarely reported. Based on this synthesis, seven development stages are proposed: (1) needs analysis and construct definition; (2) specification and blueprint development; (3) design of tasks and rubrics; (4) expert validation; (5) limited pilot testing; (6) empirical analysis; and (7) revision, final assembly, and reporting. This framework is intended to serve as a guide for educators and researchers in developing systematic and accountable maharah kalam evaluation instruments.