Detecting Groups of Test Takers Involved in Test Collusion as Unusually Large Cliques in a Graph (RR 18-01)

Test collusion (TC) is the sharing of test materials or answers to test questions (items) before or during a test. Because of the potentially large advantages for the test takers involved, TC poses a serious threat to the validity of score interpretations. The proposed approach applies graph theory methodology to response similarity analyses to identify groups involved in TC while minimizing the false-positive detection rate. The new approach is illustrated and compared with a recently published method using real and simulated data. The results of computational studies demonstrate advantages of the new approach, particularly a remarkable robustness to the multiple-comparison problem while still demonstrating good power to detect moderate to high amounts of collusion.