PREDICTING STUDENT PROFICIENCIES IN KNOWLEDGE COMPONENTS

Embodiments of the present invention allow an online course to predict proficiency of knowledge components using student responses to assessment items. In some embodiments, assessment items are tagged with metadata, such as associated knowledge component, difficulty, and depth of knowledge. Initially the author or subject matter expert supplies these values. Students then answer the assessment items. During an assessment, the student responses are automatically evaluated for correctness. Each student response receives a score between 0 and 1; where 1 corresponds to a completely correct response and 0 an incorrect response. Student proficiency can be predicted at any time using the student's up-to-date assessment item response history. Some embodiments use Samejima's Continuous Response Model (CRM), which is a special case of an Item Response Theory (IRT). Question metadata such as question difficulty, difficulty of knowledge component, impact of learning gain and retention is used to predict the most recent proficiency.