Individual differences in replicated multi-product experiments with Thurstonian mixed models for binary paired comparison data

•Accounting for replications of assessors within sensory discrimination testing.•Replications of assessors are handled with a random assessor effect in the model.•Introducing how to obtain estimates of the assessors from the suggested model.•Enabling an analysis of multiple products accounting for replicates of the assessors.•Embedding the model into a Thurstonian framework leading to d-prime estimates.