Automatic Item Generation

Automatic Item Generation (AIG: Bejar, 1993; Embretson & Yang, 2007; Gierl & Haladyna, 2013; Irvine & Kyllonen, 2002) is an assessment theory and practice that integrates cognitive/psychometric theories with digital technologies for futuristic assessments. AIG provides innovative solutions to overcome the shortcomings of the traditional approach in which human writers manually create, review, revise, and analyze individual items which is a slow, expensive, and often subjective procedure.

The essence of AIG is Computerized Item Modeling (CIM) which is an approach to program Item Model for a machine to strategically generate a large amount of high-quality digital items. AIG is a symbolic technique/theory of the initial step, the so-called “item-to-item model” transition, for the digital transformation of assessment. (Choi, 2017)

CAFA 2020