AI Summary & Key Takeaways
AI InsightFinally, it is also suggested that measurement and evaluation specialists should be employed in each school or in each local Ministry of National Education Authority at least, in order to support teachers, but if this is not possible in a close time, there must be in-service training programs on measurement and evaluation for teachers to participate in.
📌 3 Key Takeaways
Abstract
Teacher-made tests (TMT) are the most used instruments for assessment and evaluation. This study investigates the cognitive requirements, test construction errors, and item types of TMTs. Content analysis technique is used in order to analyze and classify TMT items based on TIMSS-2019 assessment framework and based on criteria that is constructed to determine test construction errors. The data is consisted of 548 items in 30 exam papers of 18 mathematics teachers from 13 distinct schools. The distribution of TIMSS-2019 cognitive demands of all TMTs indicates that there is a strong emphasis on knowing or applying cognitive domains, with a total percentage of 93. Since 83% of all questions are of multiple choice and 17% are constructed-response type, teachers mostly prefer multiple choice item type. Findings also reveal that except face validity, there are errors concerning test constructions. Consequently, it is suggested that teachers should give more care on preparing items of higher cognitive levels, on tests of mixed type items, and on tests that involve lesser construction errors for more reliable tests. Finally, it is also suggested that measurement and evaluation specialists should be employed in each school or in each local Ministry of National Education Authority at least, in order to support teachers, but if this is not possible in a close time, there must be in-service training programs on measurement and evaluation for teachers to participate in.