Institute of Technology of Cambodia

Food Technology and Chemical Engineering

SENSORY EVALUATION

Lectured by: Molika YIN (MSc)

Email: [email protected]

Academic Year: 2017-2018

 TEACHING OUTLINES
1. INTRODUCTION
2. SENSORY ATTRIBUTES
3. SENSORY PERCEPTION
4. TEST ROOM, SAMPLE PREPARATION AND
SAMPLE PRESENTATION
5. SELECTION and TRAINING of PANEL MEMBERS
6. SCALES of MEASUREMENT
7. SENSORY TEST METHODS
8. DISCRIMINATION TEST
9. DESCRIPTIVE ANALYSIS
10. AFFECTIVE TEST
 INTRODUCTION
Definition: sensory
evaluation is a scientific
method used to evoke,
measure, analyze and
interpret those
responses to products
as perceived through
the senses of sight,
smell, touch, taste and
hearing.
 INTRODUCTION
Role of sensory evaluation
In R&D, QA, QC and Marketing: to provide
insights to help guide development and
commercial strategy.
Sensory and consumer testing provide insights
into human behavior and perception.
In the early stages of product development,
consumer and sensory testing can help identify
the important sensory attributes driving
acceptability across a product category.
 INTRODUCTION
Role of sensory evaluation
It can identify sensory-based target consumer
segments, analyze competitor products and
evaluate new concepts.
From a marketing perspective, sensory and
consumer testing can inform understanding
concerning product preferences and
acceptability. It can provide the data to support
marketing claims such as best ever, new
creamier, and most preferred.
INTRODUCTION
 INTRODUCTION
How to conduct a sensory study?

1. Determine the test Objective

2. Screen the sample
3. Design the test
4. Conduct the test
5. Analyze the data
6. Interpret and report the results
SENSORY ATTRIBUTES
Appearance
Odor/ Aroma/ Fragrance
Flavor
Consistency and Texture
Noise
 Appearance

Color: visual system of light of wavelengths

400500 nm (blue), 500600 nm (green and
yellow), and 600800 nm (red)

This image cannot currently be displayed.

 Appearance
Size and shape: Length;
thickness; width; particle
size; geometric shape
(square, circular) as
examples.
 Appearance
Surface texture: The
dullness or shininess of a
surface, wet or dry, soft
or hard, crisp or tough.
 Odor/ Aroma/ Fragrance
Odor: The odor of a product is
detected when its volatiles
enter the nasal passage, and
they are perceived by the
olfactory system.
Aroma is the odor of a food
product,
Fragrance is the odor of a
perfume or cosmetic.

Olfactory System
Source: Sinauer Associates. 2001. The
Organization of the Olfactory System:
http://www.ncbi.nlm.nih.gov/books/NBK10982/
 Odor/ Aroma/ Fragrance

Odor/ aroma/ fragrance

The amount of volatiles that escape from a product is
affected by the temperature and the compounds nature.
Many odors are only released when an enzymic reaction
takes place at a freshly cut surface.
Odorous molecules must be transmitted by a gas that can
be the atmosphere, water vapor, or an industrial gas, and
the intensity of the perceived odor is determined by the
proportion of such gas that comes into contact with the
observers olfactory receptors.
 Flavor

Flavor has been defined as the sum of

perceptions resulting from stimulation of the
sense ends that are grouped together at the
entrance of the alimentary and respiratory
tracts.
 Flavor
Defined in this manner, flavor includes:
The aromatics, i.e., olfactory perceptions caused by
volatile substances released from a product in the
mouth via the posterior nares

The tastes, i.e., gustatory perceptions (salty, sweet,

sour, bitter) caused by soluble substances in the mouth.
 Flavor

Defined in this manner, flavor includes:

The chemical feeling factors that stimulate
nerve ends in the soft membranes of the
buccal and nasal cavities (astringency, spice
heat, cooling, bite, metallic flavor, umami
taste)
 Consistency and Texture

Consistency are those perceived by

sensors in the mouth other than taste
and chemical feeling.
Consistency is for non-Newtonian or
heterogeneous liquids and semisolids
for example: purees, sauces, juices,
syrup, jellies, and cosmetics.
In practice, some standardization,
consistency can be measured by
consistometers.
 Consistency and Texture

Texture can be defined in term of their:

Reaction to stress, measured as mechanical properties
(such as hardness/firmness, adhesiveness,
cohesiveness, gumminess, springiness/resilience,
viscosity) kinesthetic sense in the muscles of the hand,
fingers, tongue, jaw, or lips.
Tactile feel properties (touch), measured as geometrical
particles (grainy, gritty, crystalline, flaky) or moisture
properties (wetness, oiliness, moistness, dryness) by the
tactile nerves in the surface of the skin of the hand, lips,
or tongue.
 Noise

The noise produced during mastication of foods or

handling of fabrics or paper products is a minor, but
not negligible, sensory attribute.
It is common to measure the pitch, loudness, and
persistence of sounds produced by foods or fabrics.
The pitch and loudness of the sound contribute to the
overall sensory impression. Differences in pitch of
some rupturing foods (crispy, crunchy, brittle) provide
sensory input that is used in the assessment of
freshness/staleness.
SENSORY PERCEPTION
Human senses
 Human Senses
1. Vision
The appearance of any object is
determined by the sense of vision. Light
waves reflected by an object enter the
eye and fall on the retina. The retina
contains receptor cells, known as rods
and cones, which convert this light
energy into neural impulses that travel via
the optic nerve to the brain. Cones are
responsive to different wavelengths of
light relating to colour. Rods respond
positively to white light and relay
information concerning the lightness of
the colour. The brain interprets these
signals and we perceive the appearance
(colour, shape, size, translucency,
surface texture, etc.) of the object
 Human Senses
2. Gustation
The sense of taste involves the perception
of non-volatile substances which, when
dissolved in water, oil or saliva, are
detected by taste receptors in the taste
buds located on the surface of the tongue
and other areas of the mouth or throat.
The resulting sensations can be divided into
five different taste qualities salty, sweet,
sour, bitter and umami.
It is a myth that only certain areas of the
tongue are sensitive to particular tastes. In
fact, different areas of the tongue can be
responsive to all the taste qualities;
http://blogs.sunvalleygroup.co.za
however, some areas are more sensitive
than others.
 Human Senses
3. Olfaction
Volatile molecules are sensed by olfactory receptors
on the millions of hair-like cilia that cover the nasal
epithelium (located in the roof of the nasal cavity).
Optimal contact is obtained by moderate Inspiration
(sniffing) for 12 s. At the end of 2 s, the receptors
have adapted to the new stimulus, and one must
allow 520 s or longer for them to de-adapt before a
new sniff can produce a full-strength sensation.
The ear and eye can sense only one type of signal,
however, the nose can sense enormous
discriminating power.
Ex: a trained perfumer can identify 150-200 different
odor qualities.
 Human Senses
4. Touch
Somesthesis (tactile sense or skin feel): The skin,
including the lips, tongue and surfaces of the oral cavity,
contains many different tactile receptors that can detect
sensations related to contact/touch, e.g. force, particle
size, heat, cold, itching etc.
Kinesthesis: Nerve fibres in the muscles, tendons and
joints sense tension and relaxation in the muscles,
allowing the perception of attributes such as heaviness
and hardness.
Chemesthesis: Some chemical substances can stimulate
the trigeminal nerves situated in the skin, mouth and
nose to give hot, burning, tingling, cooling or astringent
sensations, e.g. piperine in pepper, capsaicin in chilli
pepper, carbon dioxide in fizzy drinks, coolants in
showers gel, warming compounds in muscle rubs and
tannins in wine
 Human Senses
5. Audition (hearing)
Sound is sensed by millions of tiny hair
cells in the ear that are stimulated by
the vibration of air from sound waves.
The noise created when touching or
stroking objects, e.g. fabric, gives an
indication of texture. The noise emitted
by food during eating contributes to the
perceived texture of a food, e.g.
crispness of an apple and fizz of a
carbonated drink. When consumers eat
food products, the sound waves
produced can be conducted by the air
and/ or bones in the jaw and skull.
 TEST ROOM,
SAMPLE PREPARATION
AND SAMPLE PRESENTATION
 Test Room
Should be centrally
located, easy to reach,
free of crowding,
comfortable, quiet,
temperature controlled,
free from odors and
noise.
 Test Room

To prevent subjects
from interacting.

Simple booth
 Test Room

Circular table with lazy Susan Round-table

Used for training, discussion and making decision for

descriptive analysis ballot (secret voting) development.
 Test Room

(1) Tap water;

(2) Small sink;
(3) Electrical outlet and signal
switch to panel attendant;
(4) Table covered with
odorless Formica or other
easy-to-clean surface.

Special booth features

 Walls should be off-white
shadow-free
Incandescent lighting allows
wider variation and permits the
use of colored lights, but more
heat is generated requiring
adequate cooling.
Fluorescent lighting generates
less heat and allows a choice of
whiteness.
The lights are used to mask visual
differences between samples in Panel booths showing
difference tests calling for the arrangements for lighting. (1)
subject to determine by taste (or Incandescent; (2) fluorescent; (3)
holder for sheet filters
by feel, if appropriate) which
samples are identical
Layout for medium-size
sensory evaluation
Area (Total = 704 Sq. Ft.)
 Test Room

Layout for large sensory evaluation area

 Test Room
Air Circulation, Temperature,
and Humidity

Temperature should be at 72-75

oF (22-24 oC);

Humidity: 45-50%;
Recirculated and makeup air
should pass through a bank of
activated carbon canisters that
are capable of removing all
detectable odor.
 Test Room
Construction Materials

Smooth and easy to clean

Nonabsorbent
Stainless steel, Teflon, and
Formica are the best
Off-white color
 Safety of test sample
1. Sample ingredients

The sensory professional must know the ingredients in

the products being tested in order to perform a safety
assessment, disclose information to assessors regarding
potential allergens and provide information to medical staff
in the event of an adverse reaction

2. Microbiological safe
Sample must be cleared for microbiological safety
Be stored appropriately
 Safety of test sample
3. Allergens

Assessors must be notified if the sample contains any

known allergens: gluten (wheat, oats, barley, rye,kamut,
spelt and triticale), peanuts, tree nuts, soya beans,
sesame seeds, fish, crustacean, milk, egg, sulphites
(at levels above 10 mg/kg expressed as SO2), mustard,
celery, lupin, molluscs and most products or
derivatives of these foods.
 Safety of test sample
4. Amount of product consumed
Consider short term and the long term of adverse
effects
If greater than normal amounts will be
ingested/used in one sitting, then consideration
must be given to whether the recommended daily
intake (RDI) will be exceeded and if so, a safety
assessment should be carried out.
Example: high caloric content may cause weight
gain in the long term,
Alcoholic content may cause slowed reactions
and may put assessors at risk of being over the
legal limit when driving.
Sniffing a large number of fragrances in an
alcoholic base in one sitting may cause
dizziness.
 Sample Preparation
1. Material

Materials used should be know the origin and

history of storage,
Safety and hygiene,
 Sample Preparation
2. Equipment

Glass, glazed china or stainless

steel are the most appropriate as
they are inert.
Wood should be avoided as it can
be unhygienic and can absorb food
materials which may be transferred
to other samples,
Plastic is also capable of absorbing
volatile compounds which could then
be transferred to later samples.
 Sample Preparation
3. Method preparation

Procedures should be standardized,

Use of balances, volumetrics, stopwatches, agitation
rates, temperature probes or loggers, etc.,
 Sample Presentation
1. Sample size and temperature

Sample size should be sufficient;

Generally hot food should be served at 60-66 oC;
Hot tea and coffee: 66-71 oC;
Cold beverage: 5-9 oC;
Frozen desserts: - 18 to -10 oC.
 Sample Presentation
2. Carrier

Some products, e.g. fat spreads, breakfast cereals,

seasonings and flavourings, are not readily evaluated on
their own and require an additional product to be
presented as a carrier.
Unsalted crackers, white bread, rice, pasta, milk and
bland white sauces should be considered.
For nonfood products, for example, fragrance products,
skin, hair, towel, etc, are the best carriers.
 Sample Presentation
3. Coding

Codes used for samples should be carefully recorded

and random three-digit codes are preferred.
Randomly generated by computers and calculators
or obtained using random number tables,
although certain codes should be avoided (for
example: 999, 911, )
 Sample Presentation
4. Reference samples

The purpose of the reference should be made clear to

the assessor.
The use of a reference sample should be
standardized across assessors so as not to add a
further source of variation to the experiment. The
reference may also be presented as a coded sample
to evaluate the consistency of the assessors.
SELECTION AND TRAINING OF
PANEL MEMBERS
 Recruitment
The method of recruitment will depend on the number and
type of assessors needed.
1. Advertisement
local papers, notice boards in public places, handed out in
person, posted by hand or mailed out
Adverts must be cleared, unambiguous, local employment
law,
Contact information
2. Direct recruitment
By telephone or face-to-face
3. Word of mouth/recommendation
Recommendation from existing participants can be a very
effective method
 Internal vs. external assessors/panels
Panel Advantages Disadvantages
Internal May less expensive May be more expensive
Readily to work Lose attention on sensory testing as
Flexible a priority
Need little notice May have to much product or
Well keep the secrete or confidential project knowledge
of the company Unfair or conflict
No time or less time available for
training
Not put all time and effort into test
Not available together
Often, only available for short tests

External Readily to work May be more expensive

All available together Causing trouble in group discussion
Well trained Sociable aspect could affect the
Put all time and effort into task work
Require regular supply of work for
efficient operation (less flexible)
 Screening and selection
To find out the potential or appropriate participants
However, more complex the selection criteria, the more
difficult, time-consuming and expensive.
1. Naive assessor
Untrained sensory panel: participates in very simple
sensory tests, ex: paired comparison. Screening for
sensory acuity depends on the test objective
Consumers: Consumers may be selected for
affective test; a consumer panel used to provide
affective judgments on a frequent basis.
2. Trained panel
 Trained Panel
Personality: able to work as team, experiences,
positive, good listener and communicator, committed,
flexible.
Health: good general health, physiological or health
restriction must be documented , ex: allergies, false
teeth, migraines, etc.
Sensory acuity: High sensitivity to certain stimuli
Detecting stimuli: identify specified attributes
Discriminating between stimuli: discriminate attributes
between samples
Recognizing and describing stimuli: recognize and
name stimuli
 MEASUREMENT
Nominal scales
Ordinal scales
Interval scales
Ratio scales
Hedonic scales
Face scales
 Nominal scales
Nominal scales is used in classification or naming.
Numbers are used to label, code (ex: three digit, 251), or otherwise
classify items or responses. These numbers have no real meaning.
For example Juice: 384= orange juice, 872= apple juice

Mathematics permissible for nominal scale data include frequency

counts and distributions, modes, chi-square (2), and a coefficient
of contingency.
 Ordinal scales
Ordinal scales use either numbers or words
organized from high to low, most to least,
etc., with respect to some attribute of a product set.
Ranking is one of the most commonly used types of
ordinal scale. For example: rank products from most
to least sweet or from most to least liked.
The paired-comparison test is a special use of the
rank-order test, as are the directional discrimination.
For example: which sample is sweeter.
Ordinal scales
 Interval scales
Interval scale: the distance between points on
the scale.
Panelists place the items into numbered
groups separated by a constant interval, e.g.,
1, 2, 3, 4, or 2, 4, 6, 8,
 Ratio scales
Panelists use numbers that indicate how many
times the stimulus in question is stronger (or
saltier, or more irritating) than a reference
stimulus presented earlier.
 Hedonic scales
9-points hedonic scales is generally used to
measure the acceptance/preference of
product.

Hedonic scales is simple to describe and easy

to use.
 Hedonic scales

9
8
7
6
5
4
3
2
1
 Face scales
These scales were primarily intended for use
with children and those with limited reading
and/or comprehension skills.

The facial expression may be accompanied by

a descriptive phrase and may have five, seven,
or nine categories.
Face scales
SENSORY TEST METHODS
 Sensory Test Methods
Sensory test methods are designed to answer
the following questions:
Is there a difference?
What is the nature of the difference?
Is the difference acceptable?
Sensory Test Methods (STM)

STM

Objective Subjective
By trained assessors By untrained assessors

Discriminative test Affective or consumer test

Descriptive test Acceptability

Liking
Preference
DISCRIMINATION TEST
Triangle test
Duo-trio test
Difference from control test Total different
Same-different test
A & not A test
Paired comparison
3-Alternative forced choice Attribute-specific test
Ranking test
 Discrimination Test
The most common methods employed in
sensory science,
They are used to determine if a difference (or
similarity) exists between two or more
samples,
These are often used when the samples are
considered to be confusable, i.e. their
differences are not obvious but need to be
investigated
 Triangle Test
To determine if a difference exists between two samples.
Assessors are presented with three samples and told
that two samples are the same and one is different. They
are asked to assess the samples in the order provided
and determine which sample is the odd one out. They
may also be asked to describe the difference.
Experimental design: There are six possible orders of
sample presentation. They are:
AAB BBA
ABA BAB
BAA ABB
24-30 assessors
Triangle Test
Triangle Test
Triangle Test
 Duo-trio Test
To determine if a difference exists between two samples.
Assessors are presented with three samples, two blind coded
and one labelled as a reference. They are asked to assess
the reference sample, followed by the two coded samples (in
the order provided) and determine which is the most similar
(or different) to the reference
Experimental design: There are four possible orders of
presentation in which either sample can be used as the
reference. These are:
Ref A AB
Ref A BA
Ref B AB
Ref B BA
minimum assessors are 32
Duo-trio Test
Duo-trio Test
 Difference from control test
To determine if a difference exists between one
or more samples and a control sample, and to
determine the size of the difference between the
sample(s) and the control.
Assessors are asked to assess the two samples
and determine if a difference exists between
them. They are provided with a scale to record
the magnitude of the difference. Appropriate
palate cleansers should be used after each
sample.
 Difference from control test
This test is most commonly used for quality control
where the assessors are trained to understand the
scale and the typical variation in production
samples.

Typically 2050 subjects are required to determine

the degree of difference.

When this method is integrated into a QA/QC

procedure, the number of highly trained assessors
may be as few as five.
1
2
3
4
5
6
7

* Alternative: 0 to 9 where 0 is no difference and 9 is very large difference.

 Difference from control test
The raw data are analysed using ANOVA.

Dunnetts multiple comparison test (MCT) is

used to determine which samples are
significantly different to the control. Dunnetts
test is a specialised MCT used for comparisons
against a control. Other MCTs, e.g. Fishers
LSD, can be used to determine if significant
differences exist between the test samples.
 Same-different test
To determine if a difference exists between two
samples.
Assessors are presented with a pair of samples and
asked to determine if the samples are the same or
different; they may also be asked to describe any
differences. Samples are labelled with three-digit
codes (blind coded) and should be assessed in the
order provided (left to right).
Assessors are asked to indicate how sure they are
about their decision, using a simple category scale,
e.g. very sure, sure, unsure and very unsure.
 Same-different test
Experimental design: There are four possible sample
presentations. They are:
AA BB
AB BA
30-50 assessors, can be increased as many as 200

Data analysis: The chi-squared test (2) is used to

compare sample presentations that are the same (AA
and BB) with those that are different (AB and BA)
 A and not A test
To determine if a difference exists between two
samples.
Assessors are presented with two samples, A and
not A.
Assessors are usually given as much time as
necessary to familiarize themselves with the samples.
These are then removed and the assessors are
presented with a series of individual samples, labeled
with random three digit codes, and asked to determine
if they are the same as A or not A.
Using a simple category scale, e.g. very sure, sure,
unsure and very unsure
 A and not A test
Experimental design: Usually, 1050 assessors are
trained to identify the A and not A samples. During the
test, assessors receive either:

one sample (either A or not A);

two samples (A and not A);
several samples (up to 20 samples, equal numbers of A
and not A).
 A and not A test
Data analysis: The chi-squared test (2) is
used to compare the different sample
presentations and their responses. When
calculating by hand, the 2 statistic is compared
to a statistical table.
 Paired comparison test
To determine if a difference exists between two samples
with regard to a specified attribute, e.g. sweetness,
hardness and intensity of fragrance.
Assessors are presented with two blind coded samples.
They are asked to assess the samples and determine
which of the two has the greatest intensity of a specified
attribute.
Assessors may be pre-trained on the attribute,
depending on the test objectives
 Paired comparison test
Experimental design: Samples are presented in pairs.
There are two possible orders of presentation which
should be used an equal number of times. They are:

AB

BA

A minimum assessors is 30.

Paired comparison test
 Paired comparison test
Data analysis: Determine the total number of
times each sample is selected. There are two
ways of analyzing the data. When calculating by
hand, the larger number of responses for one
sample is compared to statistical tables
 3-Alternative forced choice test
To determine if a difference exists between two
samples with regard to a specified attribute, e.g.
sweetness, hardness and intensity of fragrance.
Two samples are the same and one is different,
although the assessor is not made aware of this
fact. They are asked to assess the samples in the
order provided and determine which sample has the
highest intensity of a specified attribute.
Assessors may be pre-trained on the attribute,
depending on the test objectives.
 3-Alternative forced choice test
Experimental design: There are only three possible
orders of sample presentation. They are
AAB
ABA
BAA

minimum of 24 assessors
 Ranking test
To determine if a difference exists between three
or more samples with regard to a specified
attribute, e.g. sweetness, hardness and intensity
of fragrance.
Assessors are presented with several blind coded
samples. They are asked to assess the samples in
the order provided and place them in order of
intensity for a specified attribute. Assessors may
be pre-trained on the attribute depending on the
test objectives.
 Ranking test
as many as 810 samples for simple products
such as mineral water,
More typically, 56 samples are used when the
intensity of taste/flavour attributes is being ranked.
Critical values table for Friedman test
DESCRIPTIVE ANALYSIS
AFFECTIVE TEST
INTRODUCTION TO SPSS
PROGRAM

Statistical Package for the Social Sciences

Input variables and other
properties
Number of panelist

Liking score

View variables
Analyze the data
Prepared by: Pheakdey YUN, MSc
Email: [email protected]
2016