The Suitability of HOTS-Based Geography Questions in Terms of Validity, Reliability, and Effectiveness of Distractors

Abstract

This research is based on the importance of preparing evaluation instruments that are able to measure higher order thinking skills in a valid, reliable, and fair manner. Therefore, the purpose of this study is to analyze the quality of question items in Higher Order Thinking Skills (HOTS)-based Geography subjects that have been tested on class XI students of SMAN 2 Cikampek in the 2024/2025 academic year. The method used is descriptive with a quantitative approach, with a sample of 30 students selected randomly. The research instrument was in the form of 20 multiple choice questions that were done online through the Quizizz platform. Data analysis techniques include validity testing using product moment correlation, reliability testing using the Kuder-Richardson formula (KR-21), as well as analysis of difficulty level, differentiating power, and the effectiveness of exacerbators. The results showed that 14 out of 20 questions (70%) were valid with a high reliability value of 0.8122. Based on the level of difficulty, 65% of questions are classified as easy and 35% are classified as moderate, without any difficult category questions. Differentiating power analysis showed that 57% of the questions were categorized as very good, 21% as good, and the rest were classified as sufficient and insufficient. Meanwhile, most of the exemptions did not function optimally. These findings indicate that although most questions have met the criteria for good quality, improvements are still needed, especially in the aspect of exemptions.