Incorporating test-taking engagement into the item selection algorithm in low-stakes computerized adaptive tests

Abstract In low-stakes assessment settings, students’ performance is not only influenced by ability level but also their test-taking engagement. computerized adaptive tests (CATs), disengaged responses (e.g., rapid guesses) that fail to reflect true levels may lead the selection of less informative items and thereby contaminate item estimation procedures. To date, researchers have developed various approaches detect remove after test administration completed alleviate negative impact low engagement on scores. This study proposes an alternative method based Maximum Fisher Information (MFI) considers as a secondary latent trait select most optimal both The results post-hoc simulation studies indicated proposed could optimize improve accuracy final estimates, especially for low-ability students. Overall, showed great promise tailoring CATs Practitioners are encouraged consider incorporating into algorithm enhance validity inferences made from CATs.