ABSTRACT
This chapter reviews the history of the methods used to detect these threats, describes the nature of these threats with respect to the integrity of test scores, and defines unusual erasures and gains scores. The chapter also reviews the application of several quantitative methods to detect these threats. B. A. Jacob and S. D. Levitt provide an early, well-known attempt for using gain scores to detect teacher cheating. Cheating that involves erasures and large score gains typically occurs at the group level, such as in classrooms and schools. The chapter explores modeling and flagging criteria based on the conditional relationship between wrong-to-rights (WR) and testing and evaluations. The WR sum is frequently used as a dependent variable in erasure studies. There are several ways that the erasure data may be analyzed. For example, the WRs may be summed and analyzed at the student level, or they may be aggregated and analyzed at the classroom, school, or district level.
