<![CDATA[The use of polytomous data in test instruments enables more detailed assessment of test-takers' abilities, but group differences, such as gender, class, and ethnicity, are often overlooked. Differential Item Functioning (DIF) analysis helps determine whether these identities influence test performance. This descriptive quantitative study examines DIF in geometry tests using Holistic and Analytical Scoring Rubrics across gender, class, and ethnic groups. The study involved 102 undergraduate students from Cenderawasih University, Papua, who completed a geometry test with 10 descriptive questions. Two scoring rubrics were used: the Holistic Scoring Rubric with three categories and the Analytical Scoring Rubric with five. Results were analyzed with the difR package in the R Program. The criterion for detecting DIF is a p-value less than 0.05. Findings show that some test items exhibit DIF concerning gender, class, and ethnicity. The DIF detected for the holistic scoring rubric is items 1, 6, 7, and 10 for the gender group; items 1, 4, 5, and 8 for the class group; and items 1, 2, and 6 for the ethnic group. Meanwhile, the DIF detected for the analytic scoring rubric is item 10 for the gender group; items 9 and 10 for the class group; and item 10 for the ethnic group. However, not all DIF items are flawed; some assess fundamental skills. The Analytical Rubric demonstrated slightly higher reliability (alpha 0.903) than the Holistic Rubric (alpha 0.804). These insights support the development of more equitable and sustainable assessment practices, ensuring fairness and inclusivity in educational evaluations.]]>
Copyrights © 2026
