We have been interested in Rapid Methods for a while now: we presented our paper comparing the three methods – Napping, Flash and Free Choice profiling – at Eurosense. Since then we have seen an increase in requests for work and information amongst our client base. So what are the Rapid Methods and what is their appeal?
The main attraction is getting results faster and with less resource compared to more traditional QDA type profiling.
QDA profiling uses a group of assessors selected on the basis of their sensory acuity who are then trained in the use of rating scales (often unstructured line scales). For each product set the panel subsequently evaluates, it will collectively derive and agree a vocabulary of terms to describe all the aroma, flavour, texture and aftertaste characteristics of the products. The panel will then agree rating scale ranges for the products and spend time training to ensure that each member recognises and rates the attributes consistently. Once repeatability has been achieved the panel will move onto rate the product set, usually on replicate occasions.
Depending on the expertise of the panel, frequency of test sessions and the number and nature of the test products, a QDA type profiling project may take several days or even weeks to complete.
Free Choice Profiling was the first Rapid Method to be described. Developed by Williams and Arnold in the 1980s, Free Choice removes the need for a consensus panel vocabulary by allowing each assessor his or her own list of descriptive attributes. Each assessor then rates the samples for the attributes in his or her vocabulary using unstructured line scales. Each assessor may have a different number of attributes and this, and differences in the way they use the rating scales, are accommodated by the data analysis technique, Generalised Procrustes Analysis (GPA).
Flash Profiling was developed from Free Choice in the early 2000s. It attempts to deal with the problem of inconsistencies in line scale use by replacing the rating phase by ranking. As with Free Choice, each assessor derives their own individual list of sample descriptors but instead of then rating the samples, they are ranked for intensity. Again the data are analysed by GPA.
Napping is the most recent method to be developed and is an extension of sample sorting. The term Napping comes from the French ‘le nappe’ meaning tablecloth. The assessor is given a large surface (or large sheet of paper) and is asked to arrange the samples on the surface such that samples that are similar are placed close together and further away from those that are more different. The criteria for arranging the samples is left entirely up to the assessor. The data are collected as a series of x,y co-ordinates for each sample and each assessor. Napping data are analysed by Multiple Factor Analysis.
Fundamentally, Napping is a non-descriptive technique but is often adapted by asking assessors to note down the sensory attributes that are associated with the sample sets they form. The descriptors are then handled as a second stage of the data analysis.
Each of these methods has its unique advantages and disadvantages. I am sure you have your thoughts which we would love to hear.