Test and Scale Development and Maintenance
R. Darrell Bock, Robert D. Gibbons · 2021
This chapter covers a variety of topics related to the practical aspects of the use of item response theory (IRT) in the development and maintenance of tests and scales. It begins with an overview of item banking and calibration. The chapter then follows with a discussion of test equating and item parameter drift. Score equating is important for high-stakes testing where alternate forms of a test are used over time to reduce familiarity with the questions that might produce artificial inflation of ability estimates. Holland et al. define linkage as the transformation from a score on one test to another test. They describe three types of linkage: predicting, scale aligning, and equating. Harmonization and test linking share some features in common. Gibbons et al. use the bifactor model to provide harmonization between the two measures. Here each measure represents a subdomain and all items load on the primary dimension as well.