Assessing the Effectiveness of CAT and MCAT in Measuring Students' Understanding in Technology Education

This research aims to do two things. First, it aims to make a computer adaption test (CAT) for Islamic subjects for the santri in ABC Pesantren. Second, it aims to see the results of the CAT test for Islamic subjects in the production of Learning Medical Records. The records is for diagnosing students' abilities. This research consists of two parts using the research and development (R&D) approach: 1). We collected the data by observing, documenting, and interviewing. We analyzed the data using the concept of reliability among the raters/respondents. We used the generalizability coefficient in Genova 3.1. We analyzed the results using numbers. They describe data from RMP technology education. The research shows that: (1) the CAT can be made based on user needs. It's Internet-based, user-friendly, interactive, highly secure, and easily accessible. (2) The CAT can recognize three different users: school management, teachers, and students. (3) The expert evaluation's quantitative data has high consistency: 0.76. It comes from single, small group, and large group outcomes. (4) The CAT can accurately estimate students' abilities and record learning records. The software uses the triangle decision tree method. It can pick test items and measure ability well. This is clear from the correlation (r) between ability and the software's CAT scores.