ABSTRACT
The rapid development of technology provides both opportunities and challenges in the design, development, implementation, and use of assessments. Technology use is prevalent in assessment programs, from the design of technology-based items to the design of automated scoring, providing opportunities to enhance the validity of score meaning and test use. The use of educational data mining and machine learning allow for the measurement of competencies that have been difficult to measure on a large scale, including written and spoken language. An overarching goal of the use of data mining and machine learning in test design is to improve construct representation by more accurately capturing and representing examinees’ capabilities while minimizing construct-irrelevant variance. Educational data mining and machine learning allow for capturing salient features of the processes and actions used by examinees. This chapter addresses validity and fairness issues for technology-based assessments.
