|
PANEL MANAGEMENT AND CHARACTERISTICS
Standard fields are suggested, others can be inserted: availability dates, sensory timetables, ...
Codification links between assessor # and assessor name are used to print automatically assessor names in results reports, if need be.
PANEL TRAINING
It is possible to define training ranges on olfactory, gustatory, and to ask qualitative or/and quantitative recognition.
Panellists can enter data in booths during self-service sessions in order to get furthermore training results on assessor sensitivity.
PANEL EVALUATION
On the basis of discrimination test
Panel evaluation is suggested according to sensory tests (ex. triangle test), product groups (ex. FRA group),
or also assessors (by assessor or for the whole panel).
CONCLUSIONPORTIER assessor is maybe more sensitive to identify differences between two products,
but with his low attendance index his final result is lower than DUPONT, who is coming
more regularly, and possess perhaps a higher interest in sensory evaluation.
On the basis of profiles tests
 
First graphs compare each assessor results with the panel results.
Then, individual performance coefficient can be displayed.
To use these index, for the repeatability, it is noticeable, the assessor the best,
is the one with the index the lowest. An index higher than 1, means a non repeatable assessor.
For discrimination also, the index the lowest, means the assessor the best to discriminate products.
Other results are available on consensus and homogeneity.
ConclusionOxidised and sweet notes do not allow a good repeatability with high index, PERTY and QUESNES
are not discriminative for the floral note, DURANT is a very good assessor for the ripe note
(low repeatability and discrimination index), QUESNES is also an excellent expert for the mushroom note.
Similar results are also available for the whole panel.
If we are discussing about results, for discrimination, the lowest is the index
the highest is the panel ability to discriminate the products.
For the homogeneity, the lowest is the value, which is a probability, the lowest is the homogeneity
of the group, in other terms, the products are assessed differently by the experts.
For the consensus, the highest is the value, the highest is the panel consensus, this last value
is based on a multidimensional analysis of the means by assessor and by product.
ConclusionFor discrimination, the panel is better than each individual panellist value,
but with a quite high consensus panel value, and so, we can observe the interest of the use
of a panel for the assessement of products.
|
