Abstract Choosing a consumer test methodology is critical for any company sensory scientist. However, we lack tools for comparing the outcome of consumer tests. Here we intended to develop a new comparison criterion based on a concept borrowed from metrology: robustness. We tested this concept with hedonic ratings data obtained from a central location test (CLT) on one hand and from a home use test (HUT) on the other hand. Two fermented milk beverages were tested and both tests were conducted with 240 subjects. The beverages differed in fat and sugar content. The comparison of the CLT and HUT data sets (both in monadic sequential) indicates that the method has a significant effect on the degree of liking. Indeed, the liking scores are lower with CLT than with HUT. Still, both tests lead to the same conclusion that the product with the higher fat and sugar content is significantly more liked. Besides, we developed three graphic representations that illustrate the results stability faced with the decrease in the panel size. These graphs show that CLT yields more robust results than HUT. Additionally the leading role of the panel size is emphasised for both tests. Our data show that the results from the two methods may change as soon as 20 subjects are removed from our original data set. These results that were not unexpected illustrate a fact which is often underestimated in practice.