KESAHAN DAN KEBOLEHPERCAYAAN INSTRUMEN RUBRIK ANALITIKAL BAHASA ARAB BAGI TAHAP PENGUASAAN MURID SEKOLAH KEBANGSAAN

Abstract

This study examines the validity and reliability of the items in the Arabic Language Analytical Rubric Level One (BATS). The rubric was developed based on qualitative data obtained through purposive sampling involving outstanding teachers in the teaching and learning of Arabic at primary schools. The validity and reliability of this analytical rubric were analyzed using the Multi-Facet Rasch Model, assisted by Facet software. The BATS Analytical Rubric comprised 99 items (48 under the Language Skills dimension and 51 under the Exemplary dimension). The Exemplary dimension items and the Language Skills dimension items were assessed by six and seven experts respectively. Content validity was tested by examining the polarity of the items, while reliability was assessed by examining item functionality in terms of reliability and the item-person separation index. All items demonstrated positive polarity values (Language Skills: PMC between 0.52 - 0.84; Exemplary: PMC between 0.73 - 0.83). The reliability values for both items and individuals across the two dimensions were at a very high level (>0.9). The person separation index for the Language Skills dimension was 5.64, and 5.72 for the Exemplary dimension. The item separation index for the Language Skills dimension was 3.48, while the Exemplary dimension was 4.08. Based on these findings, the BATS Analytical Rubric is considered to have a high level of validity and reliability, making it suitable for use as a guide for teachers in determining students' mastery levels in Level One Arabic Classroom Assessment at national schools. In conclusion, the reliability analysis and construct validity have contributed to developing a high-quality and meaningful BATS Analytical Rubric.