Item analysis of teachnological pedagogical content knowledge (TPACK) in pre-service chemistry teachers using the Rasch Model application

This research aims to analyze TPACK items using the Rasch model in terms of validity, reliability, item difficulty and bias items. The method used is descriptive quantitative with 34 pre-service chemistry teachers in semester VI consisting of 26 female students and 8 students from Sebelas Maret University (UNS). The instrument used in this study was the matter of Teachnological Pedagogical Content Knowledge (TPACK) which was adapted from the PPG Teachnological Pedagogical Content Knowledge (TPACK), which consisted of 30 multiple choice questions. The results showed that: 1) the validity of the suitability level of items was 27 questions that were fit and 3 questions that were not fit. 2) the results of raw variance data of 33.5% indicate that the requirements for Raw variance explained by measures of at least 20% can be met and the variance that cannot be explained by the instrument of 8.2% also meet the criteria ie, not to exceed 15%. 3) Reliability of pre-service chemistry teachers is 0.52 in the weak category, item reliability is 0.90 in the good category and reliability between pre-service chemistry teachers and items 0.64 in the sufficient category. 4) The level of difficulty of the questions is very difficult, difficult, easy, very easy and is dominated by questions that are categorized as difficult. 5) there are no questions that have a probability of less than 5% so there are no biased items. In conclusion, TPACK item analysis using the Rasch model is valid, reliable, item difficulty is very good and there are no biased items. So the questions that have been analyzed can be used to measure TPACK on pre-service chemistry teachers.