<![CDATA[ This study aims to analyze the item characteristics of the mathematics problem-solving ability test instrument using the Classical Test Theory model. The data collection was based on the results of the test documentation as many as 359 students with dichotomous data. Qualitative validation analysis by experts uses the panel method, and quantitative analysis uses the Aiken index and Content Validity Ratio (CVR), while quantitative validation uses biserial point correlation. The test reliability index used the Alpha-Cronbach method. The results of qualitative validation show that all items correspond to the indicators of solving mathematical problems, but technically writing a few items needs improvement. Quantitative analysis using the Aiken index and CVR shows all items are valid with good validity. The reliability of the test is stable, with a coefficient of 0.83. Then, the test instrument consists of items with very difficult, difficult, and easy levels. All items were able to distinguish the test taker's ability and the effectiveness of the distractor to function correctly. Keywords: classical test theory, biserial point correlation, validity, reliability. DOI: http://dx.doi.org/10.23960/jpmipa/v22i1.pp98-111 ]]>
Copyrights © 2021
