ASSESSING THE VALIDITY AND RELIABILITY OF SCIENCE MULTIPLE CHOICE TEST USING RASCH DICHOTOMOUS MEASUREMENT MODEL

Multiple choice tests are widely applied to assess students’ knowledge in science education. This study aimed at assessing the validity and reliability of Science Multiple-choice Test Malaysia. The items for this test were formulated by researcher together with a panel teachers head department close reference Secondary School Standard Curriculum (KSSM) syllabus. consists 50 multiple-choice four options. Rasch measurement model was adopted evaluate quality terms analysis, item polarity analysis (PTMEA-CORR), fit Principal Component Analysis Residuals (PCAR). performed using Cronbach’s Alpha, results separation index respectively indicated good level items. In order improve test, two negatively worded (Q39 Q40) removed. Lastly, PCAR showed unexplained variance 1st contrast (5.4%) found be well controlled below ceiling value one-third explained (18.7%). However, positive disattenuated correlations indicate no evidence presence secondary dimension. Keywords: model, validity, subject, school standard curriculum