ABSTRACT
The use of examinee response process data to validate score meaning in educational testing has evolved from studies using protocol analysis to more-recent studies that evaluate log files to infer the strategies and processes examinees use when solving tasks. Methodological advances have also been made within individual procedures, such as protocol analysis, to ensure the validity of the information obtained about students' cognitive processing. Beginning in the 1970s both cognitive psychologists and psychometricians were studying the cognitive processes, knowledge, and strategies that underlie item performance. It was not enough to know whether an examinee responded correctly or incorrectly, researchers wanted to understand what cognitive processes and strategies were used by examinees when solving items and what item features, and consequently processes and strategies, affected item difficulty. In his seminal chapter on validity, Messick (1989) discussed approaches that were emerging for studying processes underlying item performance as well as linking these processes with item difficulty, including protocol analysis, computer modeling, analysis of reasons, analysis of eye movements, and analysis of systematic errors. Some of these methods for studying response processes are discussed in the chapters in this section.
