EXPLORING CONFIDENCE ACCURACY AND ITEM DIFFICULTY IN CHANGING MULTIPLE-CHOICE ANSWERS OF SCIENTIFIC REASONING TEST

Purpose – Researchers discovered that when students were given the opportunity to change their answers, a majority changed theirresponses from incorrect correct, and this often increased overall test results. What prompts modify theiranswers? This study aims examine modification of scientific reasoning test, with additional exploration on confidence accuracyand its relation item difficulty. Methodology –A set pre-test post-test experiments which included 20 items confidencejudgement each used. The instruments assessed for validity by analysing psychometric properties using three-parameter (3PL) Item Response Theory, was carried out in R studio. randomly administered 205 Indonesian undergraduate background science education related major. accuracy determined categorising correct or answers questions based level confidence. Findings –The results revealed responses modified more frequently than incorrect,resulting significant gain score although these modifications not shown be connected theitem’s difficulty level. Even though also significantly, it observed repeatedlyresponded overconfidence even after sitting same three weeks, could indicate lack metacognitive ability. findings serve spur educators begin actively engaging training teaching learning activities as result occurs among examinations. Significance –This provides further substantiation field cognitive science; trend has been observed. It contributes uncovering true ability performing such tests through repeated attempts.