A comparison of the aromas of six coffees characterised by conventional profiling, free-choice profiling and similarity scaling methods

Abstract

A selected panel, trained in sensory procedures, used a previously developed vocabulary to profile the aromas of six ground coffees. The same coffees were also examined by a different panel of assessors not familiar with coffees, using free-choice profiling and, on a separate occasion, similarity scaling procedures. The data from the three experiments were independently subjected to the appropriate statistical techniques to produce two-dimensional diagrams showing the interrelationships of the samples. All three procedures produced similar information both in respect to the way the coffees grouped and differed from one another. The paper concludes by discussing the assessor variation associated with each of the procedures and shows that free choice profiling compares very favourably with the other two.