ABSTRACT
A central issue in higher education today is the assessment of what students learn during college. Increasingly, institutions are being asked to provide evidence of student learning, especially in broad skills such as critical thinking, communication, and problem solving (Ewell, 2009; Shavelson, 2009). Institutions are turning to assessments such as the Collegiate Learning Assessment and the Measure of Academic Proficiency and Progress (now known as the ETS Proficiency Profile) to gather data that can be used to measure their effectiveness and help improve teaching and learning at the program and institution levels (Shavelson, 2008). To inform such decisions, the data collected must be reliable. This chapter describes a statistical theory for evaluating the reliability of behavioral measurements such as scores on learning assessments.
