Proficiency testing in sensory analysis is an important step towards demonstrating that results from one sensory panel are consistent with the results of other sensory panels. The uniqueness of sensory analysis poses some specific problems for measuring the proficiency of the human instrument (panel). Individuals within a panel and people from different cultures may have different thresholds of perception, and product experience of the panel may lead to differences in the ability to discriminate among samples. Such factors make the job of the sensory scientist and statistician more challenging, because of the difficulties in defining the expected level of performance. As part of an EU supported project, ProfiSens, 12 panels undertook descriptive profiling on six samples of red wine. Four were designated 'validation' panels, whose data were used firstly to establish the expected profile results, and secondly to set the performance criteria that a trained sensory panel would be expected to achieve. Four key measures of a panel's performance were investigated: (1) the number of significant sensory dimensions identified by using generalised Procrustes analysis; (2) the number of pairs of samples that a panel found to be different at a specified level of significance; (3) how well the 'sensory map' of a panel agreed with the 'expected sensory map'; and (4) how well the assessors agreed with each other within a panel, and with the consensus configuration. For each of these criteria, an 'expected result' was considered, as was an overall measure of performance. The data from the remaining eight panels were analysed, and the level of performance recorded for each of the stated criteria. Results indicated differing levels of performance. The study also revealed a number of key issues that need to be addressed to successfully run proficiency testing for descriptive profile analysis. A simpler performance scheme is proposed to address issues related to attaching arbitrary weightings to each of the performance criteria and to address potential problems associated with combining different measurement criteria into a single performance score. While some issues still need addressing, this project has made significant contributions to proficiency testing for sensory analysis. © 2002 Elsevier Science Ltd. All rights reserved.
[1]
J. Gower.
Generalized procrustes analysis
,
1975
.
[2]
R. Forthofer,et al.
Rank Correlation Methods
,
1981
.
[3]
Frank Rossi,et al.
Assessing sensory panelist performance using repeatability and reproducibility measures
,
2001
.
[4]
P. Robert,et al.
A Unifying Tool for Linear Multivariate Statistical Methods: The RV‐Coefficient
,
1976
.
[5]
Jean A. McEwan,et al.
Evaluation of an International Ring Trial for Sensory Profiling of Hard Cheese 1 Contributions from
,
1998
.
[6]
Tormod Næs,et al.
Detecting individual differences among assessors and differences among replicates in sensory profiling
,
1998
.
[7]
Jérôme Pagès,et al.
Inter-laboratory comparison of sensory profiles: methodology and results
,
2001
.
[8]
M. O'Mahony.
Sensory Evaluation of Food: Statistical Methods and Procedures
,
1986
.
[9]
M. Thompson,et al.
The international harmonized protocol for the proficiency testing of (chemical) analytical laboratories (Technical Report)
,
1993
.
[10]
Tormod Næs,et al.
Analysis of Variance for Sensory Data
,
1997
.
[11]
A European sensory and consumer study—A case study on coffee
,
1998
.
[12]
W. J. Conover,et al.
Practical Nonparametric Statistics
,
1972
.