Equivalence bounds for sensory applications based on consumer product evaluation
Equivalence testing is widely used in sensory science during all stages of product development and production. It is important to understand product differences and variations, some differences will go undetected by the consumer or will be considered “close enough” and therefore acceptable to the consumer. Rousseau (2015) demonstrated a strategy for calculating a consumer-relevant equivalence bounds using d-prime (dˈ) and tau (τ). However, all published studies using this strategy thus far are based on simulated data. The present study used four different grades of upholstery foam, a non-fatiguing product, to collect same-different and tetrad results in order to determine the equivalence bounds. 2AC preference results were also collected but not considered in equivalence bounds calculation. 292 assessors (127 male) aged 18+ (mean 42 years old) participated in tactile evaluation. Assessors moved between pre-set stations of which only a selection of orders were presented each sitting, the orders were rotated between sittings according to the design. Each assessor evaluated one possible order for each product pair (six total) in the tetrad test. In the same-different test, one same or different pair for each product pair (total of three same, three different) were evaluated by each assessor.
Results show that based on dˈ values the same-different method appears more sensitive compared to the tetrad test, contrary to the expectation that dˈ values would not differ between methods. The same-different test was used to calculate the tau which acts as an indicator of discernible differences. However, since results differed between methods, conclusions of equivalence bounds did not translate between same-different and tetrad results. Based on preliminary analysis, questions are raised whether dˈ can be used as a universal comparison tool between methods
King, S. K., Findlay, C. J., & Castura, J. C. (2017). Equivalence bounds for sensory applications based on consumer product evaluation. 12th Pangborn Sensory Science Symposium. 20-24 August. Providence, RI, USA.