Validity and reliability of students' mathematical communication instruments in class VII middle school statistics material

This research aims to reveal content validity, construct validity, and reliability of mathematical communication ability instruments. This research involved 100 junior high school students. The instrument has been validated, contains 12 multiple-choice questions and three describing mathematical communication skills, which use the V Aiken index to measure content validity. Meanwhile, construct validity was proved using factor analysis using the Kaiser Meyer-Olkin (KMO) test, and Cronbach's Alpha test for reliability estimation by R Studio. The level of difficulty, different strengths, alternative answers, and final conclusions are calculated using Anbuso version 8.0. The research results show that 1) The instrument meets the content validity criteria from the material, construction and linguistic aspects based on the satisfaction of 4 expert validators, proven by calculating the V Aiken index for all valid question items. 2) All items meet the construct validity criteria. 3) The estimated reliability for all types of questions is 0.812 for multiple choice questions with an SEM of 2.5 and 0.868 for essay questions with an SEM of 2.1, so the instrument is reliable overall. 4) The distinguishing power of the good category is 87%, while the quite good category is 13%. 5) The difficulty level for the difficult category is 7%, and the medium category is 93%. 6) All alternative answers work well. 7) Final conclusion: 93% of the 16 questions on the mathematical communication ability instrument are acceptable, and 7% need to be changed. 7% (1 question) are questions with indicators connecting diagrams, graphs, tables to mathematical ideas.