MISCONCEPTION ANALYSIS FOR INTEGER NUMBERS BY USING RASCH MODEL

Abstract: Diagnostic Test for Operations Subtraction of Negative Integer Number (UDOPNIN) is constructed to identify misconceptions among form one students on subtraction operations involving negative integer numbers. Distractor analysis in Option Probability Curves (OPC) through RASCH model in this study was to identify types of misconceptions among form one students based on students’ ability. The survey research method is used in this study to determine the achievement of the objectives of the study. A sample of 572 Form one students from 11 secondary schools in Penang state involved in this study, consisted of 298 boys and 274 girls. The analysis validity and reliability of UDOPNIN instrument by RASCH model showed standardised residual variance (PCA) is 70.4%, the PTMEA CORR is between 0.12 to 0.68 provide the item can contribute to identify misconceptions of the respondents, 38 items (95%) is fulfil the requirement in MNSQ infit and 35 items (87.5%) is fulfil the requirement in MNSQ outfit, the result of t-test in Differential Item Functioning (DIF) for 40 items is between −2.0 to +2.0, and the value of Content Validity Index (CVI) for UDOPNIN instrument is 0.80. The OPC of 40 items in UDOPNIN instrument showed among the students’ misconceptions on subtraction operations of negative integer numbers, majority of students who had problems of symbol misconceptions was because of pseudo-conception accommodation. The result of Winsteps version 3.57.2 showed that the item reliability index is 0.99 and the respondent reliability Khong Wei Seng Lim Hooi Lian 106 index is 0.90, which showed that the UDOPNIN instrument and the result of study are reliable and valid.