Three Modeling Applications to Promote Automatic Item Generation for Examinations in Dentistry

Area of Science:

• Dental Education
• Medical Informatics
• Assessment Science

Background:

• Traditional test item development for dentistry is resource-intensive, relying on individual content experts to create limited numbers of items.
• The increasing demand for dental examinations necessitates more efficient and scalable item creation methods.

Purpose of the Study:

• To describe and illustrate systematic approaches for generating large volumes of test items for dentistry examinations.
• To showcase Automatic Item Generation (AIG) as a method to enhance item production efficiency.

Main Methods:

• Utilized three modeling approaches for Automatic Item Generation (AIG): item cloning, cognitive modeling, and image-anchored modeling.
• Integrated domain expertise of content specialists with computer technology to generate multiple-choice test items.
• Combined expertise of two content specialists with AIG technology.

Main Results:

• Generated a total of 5,467 new test items.
• Demonstrated AIG's capability in item creation through content substitution, cognitive response modeling, and image-linked item generation.
• Successfully produced a large volume of items efficiently.

Conclusions:

• Automatic Item Generation (AIG) presents a viable and scalable solution for meeting the demand for dental test items.
• The described AIG methods are adaptable for various item types and can be applied beyond dentistry.
• Further research into AIG applications in dental education is warranted.