Item Selection Rules for Content Adaptive Progress Testing

Abstract

Content Adaptive Progress Testing (CAPT) introduces a new approach to longitudinal testing of candidates who take progress tests as part of their educational programme. CAPT assessments adapt between subsequent assessments tailoring items on an individual level to assist each candidate in demonstrating knowledge across many topics which are all mapped to the overall course learning outcomes. Some of the key benefits include a truly personalised learning experience, with individual feedback throughout while also tracking each candidate’s progress against a single long-term goal of demonstrating attainment across all required topics. For the learning experience to be personalised, there needs to be automatic software selection of different items for each candidate based on their performance across a sequence of assessments. The advantages, disadvantages and practical implementation of different item selection rules are discussed.  There is a need to ensure no item is repeated to the same candidate in any test; no two items in the same test belong to the same topic for the same candidate; a minimum number of topics per broader area are tested in every assessment, and selection of incomplete topics is prioritised over complete topics. These rules need to be clear and transparent to motivate appropriate learning.