ABSTRACT
As long as standardized tests have been employed, attempts have been made by subgroups of the examinee population to gain unfair advantages over others. Classically, cheating takes on three incarnations “(a) taking, giving, or receiving information from others; (b) using forbidden materials or information; and (c) circumventing the assessment process” (Cizek, 1999, p. 39). Contemporary testing agencies are faced with much more complex, sophisticated and vastly more resilient attempts to gain unfair advantages. Therefore, as part of the initiative to maintain a valid and reliable testing product, Test Security Offices (TSOs) are prepared to investigate examination administrations through a variety of techniques. If there is reasonable evidence that improper testing behavior may have occurred between two students, they are often investigated for cheating by comparing student response strings. Comparisons of this type are able to test for systematic synchronicity in answers between plausible sources and subjects (cheaters). It should be noted that, historically, cheating investigations are launched primarily when violations are reported by external agents (such as proctors or anonymous tips) or when statistical flags in post administration screenings are elevated (e.g., a large score increase on the examinee’s performance compared to those of previous administrations). What characterizes these investigations is that they are often limited and address a one-dimensional conceptualization of what cheating entails. To the knowledge of the authors, no testing program routinely screens responses of an entire exam administration.
