A Statistical Model to Assess Cognitive Skills

Castle Rock Research provides quality curriculum-based educational resources to students, parents and educators, both online and in print. Assessment of learned knowledge needs to focus on two distinct aspects: knowledge retention and associated cognitive skills. Knowledge retention not only refers to the ability to recall learned facts but also the ability to understand the relationship between these facts (ie to understand the structure of the learned domain knowledge). Traditional test types, such as multiple choice items, typically only test a student’s ability to recognize learned facts in isolation and in the same form in which they were learned. In contrast, this project applies machine-learning techniques and involves the design of test items which are aimed at testing a student’s deep understanding of a domain, namely the ability to recall facts in novel contexts (thus requiring the student to develop a domain-independent representation) and the ability to represent the structure of the acquired knowledge (thus requiring the student to develop an understanding at different levels of abstraction).

Faculty Supervisor:

Dr. Irene Cheng

Student:

Christopher Kerr

Partner:

Castle Rock Research Corp.

Discipline:

Computer science

Sector:

Education

University:

University of Alberta