Investigation Of Test Performance Over Repeated Sessions Using Signal Detection Theory: Comparison Of Three Nonattribute-Specified Difference Tests 2-Afcr, A-Not A And 2-Afc

Abstract

To investigate more flexible methods for measuring overall sensory differences, the performance of three nonattribute-specified difference test methods was compared using Signal Detection Theory. A-Not A, 2-AFC, and 2-AFCR tests were performed with experienced subjects over repeated sessions. Learning effects were investigated to determine how much training would be needed before subjects could perform these tests consistently. Results in d' showed that in early sessions, the 2-AFC and 2-AFCR performed better (higher d') than A-Not A and that after two sessions, the test performance stabilized on that level. Learning effects for the A-Not A test were poorer in the earlier sessions, with performance levels increasing, even after six sessions. In business situations where the tested products frequently vary, the use of 2-AFC and 2-AFCR would be recommended because they are more sensitive in the earlier sessions, requiring less training time than A-Not A. PRACTICAL APPLICATIONS: Sensory difference tests are important tools for business decision making. Food companies may save a lot of money by employing more effective methodologies. For sensory difference tests to be used accurately and efficiently, test performances need to be investigated with consideration for food discrimination constraints and objectives. This article explored the performance of nonattribute-specified 2-AFC and 2-AFCR tests over several sessions in comparison to the results of a standard signal detection measure, A-Not A, to use the differences between these methods as a basis for inference concerning the cognitive strategies and relative test effectiveness of nonattribute-specified 2-AFC and 2-AFCR for food discrimination, in practical situations when both the reference product and test product(s) change frequently. This article suggested the applicability of a nonattribute-specified 2-AFCR procedure for discriminating small differences without much prior training. © 2011 Wiley Periodicals, Inc.