Limitations of the Score‐Difference Method in Detecting Cheating in Recognition Test Situations

An examination is made of a score‐difference model for the detection of cheating based on the difference between two scores for an examinee: one based on the appropriate scoring key and another based on an alternative, inappropriate key. The normal approximation to the binomial distribution model is used to predict characteristics of chance test score distributions and difference score distributions. Results show that students' difference scores can vary widely by chance alone, even for short tests. It is argued that the score‐difference method could falsely accuse many students as cheaters. It is suggested that the score‐difference method is seriously flawed and has little to recommend it.