ABSTRACT
A model for student knowledge diagnosis through adaptive testing is presented in [14]. Permutational multiple choice question tests have been used for assessing high-level thinking [12]. Tests are designed to assess critical thinking applying Bloom's Taxonomy. Such tests contain stems asking students to identify the correct outcome of a given circumstance, map the relationship between two items into a different context, respond to what is missing or needs to be changed within a provided scenario, and evaluate the proposed solution based upon criteria provided. Students' conceptual thinking can be assessed by presenting them with tests where all the correct answers should be chosen and/or answers require integration of several components or approaches [3] and [15]. A test where the set of putative answers may contain several correct answers and several wrong answers allows for very detailed feedback since it provides increased accuracy, reliability, and usability. Each one of the unique answer combinations to every question implies a different level of knowledge about the topic. There is no restriction on the number of correct and wrong alternatives, which prevents students from 'intelligent' guessing about the number of correct answers following each question.
