How much sensory panel data do we need?

It is argued that using the Benjamini-Hochberg procedure (Benjamini & Hochberg, 1995) is a good way of dealing with the problem of multiple testing, since it provides adequate protection against this whilst preserving power of the individual product comparison tests. Regarding the latter, it compares very favorably with the well-known Bonferroni and Tukey methods. It is argued that the difference between products which can be detected (the delta value) should be corrected for multiple testing. It is explained how the procedure is implemented in sample size and power calculation. The methodology is extended to consumer studies and to incomplete designs. The percentage of effective pairs is introduced, allowing for a direct comparison with a full sequential monadic design. Two examples of sensory panel studies are provided.