A method devised by Gordon and Cochetti in the 1980s. All responses within each channel are ranked from largest to smallest; ranks assigned to comparison questions are given positive values, while those to relevant questions receive negative values. For example, if a test had three each of relevant and comparison questions, and the magnitude of the responses in a given channel resulted in an order of R3, R1, C1, R2, C3, and C2, their values would be designated as -6, -5, +4, -3, +2, and +1, respectively. This method is repeated for all channels in all tests and then summed for a grand total. Thresholds suggested by Gordon and Cochetti were two points per relevant question per test, and a minimum of two tests. Because of the ranking approach, this scoring system may be limited to single-issue testing situations. Additionally, some of the diagnostic criteria and transformation procedures have not been shown to be empirically supported. See: Gordon (1999); Gordon & Cochetti (1987); Gordon, Mohamed, Faro, Platek, Ahmad & Williams (2005); Krapohl, Gordon & Lombardi (2008); Nelson & Handler (2011).
