Sensometrics
Lecture statistics in sensory science
Sensometrics; Data relationships and multivariate applications:
After the lecture the student is expected to:
Know the basic principles of sensometrics
Understands Principle Component Analysis (PCA) and is able to interpret the results.
Be able to read and interpret biplots
Know the principals of mixed models as explained in this lecture, including post-hoc analysis
(LSD).
Is able to interpret the SPSS and R output of Mixed Models including post-hoc analysis (LSD).
Training and panelselection
These clips relate to those on descriptive analysis methods (see week 1).
After watching the knowledge clips about Training and selection panel, the student is expected to:
Know how to select assessors for sensory panels for different purposes.
Understand the role of the panel leader for the success of sensory projects.
Know the different manners of attribute generation.
Know and be able to apply the methods of consensus training and individual vocabularies.
Reading material Ch 7, 10.3-10.5
Clip | Training and selection panel - Selection and orientation
Why select sensory subjects? Needs people that are able to perceive and understand the sensory
properties in food products – they need to name and explain the sensory characteristics and rate
their intensity. Have a sensory ability that is above average and can be trained.
Selection procedure:
Recruitment group of inexperienced people
Internal or external
Inside the company - familiar with the products, but maybe not so creative with
attribute generation. Bias because of information from a previous test.
Outside the company – what age (20-55)? Recruit twice the number of panelist you
need in the end, because a few people will not pass the selection test.
Basis selection
Personality - team-player, positive, good communicator, committed and flexible +
cosmopolitan preferences (not picky eater and open to try new things)
, Health – general good health, documentation of any health restrictions
Availability
Test for sensory ability – basic tastes, taste intensity ranking, smell recognition, smell
memory, naming/creativity, color blindness – fast taste/smell test: taste strips, UPSIT
(smell identification test)
Task-oriented selection
General orientation – test protocols, discrimination tests, intensity scaling
Training for a descriptive panel
Attribute specific training, reference materials training, training on intensity scale
(repeatability)
It does not stop here:
Clip | Training and selection panel - The panel leader
Panel leader:
Leading the panel
Direct contact point
Monitoring agreement within the panel, performance of the panel and keeping the panel
motivated
Communicating with the researchers and developers
Request of the sensory test
Types or roles:
Passive facilitator – when the panelist use their own vocabulary, e.g. QDA.
Directive leader – when the sensory methodology has fixed and standardized vocabulary that
needs to be taught to the panelists, e.g. spectrum.
Characteristics – non-judgmental, ability to motivate, sensitive and assertive, communication skills,
diplomatic.
Responsibilities of the panel leader for the rules for the panelists:
No eating, drinking or smoking
No perfume, lipstick or nail polish
Indicate if having cold, feeling tired
Punctuality
Follow instructions
, Panel maintenance and motivation:
Performance monitoring
Sequential analysis – long term performance
Feedback on performance (individual + panel)
Motivation
Test frequency/difficulty – so that the panelist does not lose focus and maintains their
performance.
Reward and attention
Feedback
Team building
Planning sensory tests:
The test protocol
Aim of the study
Handling and preparation of samples
Expected sensory variation
Setting limits for including samples
Experimental sample design
Pre-trials
Select sensory test
Select products
Clip | Training and selection panel - Training a sensory panel
Attributes for descriptive analysis:
Sensory attributes
When you want to do descriptive analysis – use an analytical sensory panels to profile the
products
Create vocabulary for a certain product set
Start of training phase
Consensus training
1. Provide a range of products
2. Let the panelists describe the differences that the perceive individually
3. Panel leader makes a list of all attributes and definitions that were listed
4. Reduce the number of attributes in a groupdiscussion
5. Agreement on terms and definitions between panelists
Ballot training
1. Provide a range of products
2. Let the panelists describe the differences that the perceive individually
3. Provide published word list (lexicons) to the panelists
4. The panel make a suitable list of attributes from the lexicons
5. Agreement on terms and definitions between panelists