ABSTRACT
This chapter considers some implications of technology for score comparability and offers a useful framework for future investigation as both item formats and varieties of interactive devices evolve. It offers a comprehensive and informative history and overview of diagnostic classification models (DCMs), especially the log-linear cognitive diagnosis model (LCDM), which subsumes many earlier DCMs as special cases. DCMs will be most useful when applied to tests designed from the outset to differentiate among specific student misconceptions or skill profiles. The chapter shows the various ways in which computer-based testing (CBT) has spurred new developments in IRT and addresses a range of concerns beyond the obvious requirement for item-selection algorithms for computerized adaptive testing (CAT). It discusses the trajectories of technical developments in CAT versus automatic essay scoring (AES) and proposes that progress was less rapid in AES in part because much of the early research and development work was proprietary.
