Sensory Measurements of Texture

1
Sensory Measurements of Texture

2
Texture is the sensory and functional
manifestation of the structural and mechanical
properties of foods, detected through the senses
of vision, hearing, touch, and kinesthetics. A
lina Surmacka Szczesniak (1963) Texture is a
sensory property. It is not the force needed to
push a needle-like probe a certain distance into
an apple, or the work necessary to disintegrate a
handful of peas in a shear cell. Texture makes
sense only when viewed as a sensory property or
how a food feels in the mouth Alina
Surmacka Szczesniak (1998)
3
Sensory methods are the ultimate method of
calibrating instrumental methods of texture
measurement . . . Sensory evaluation offers the
opportunity to obtain a complete analysis of the
textural properties of a food as perceived by the
human senses. A number of processes occur while
food is being masticated, including deformation,
flow, comminution, mixing and hydration with
saliva, and sometimes changes in temperature,
size, shape, and surface roughness of the food
particles. All of these are recorded with great
sensitivity by the human senses, but many of them
are difficult to measure by objective
methods. Malcom Bourne (1982) Man is
nothing but a bundle of sensations. Protagoras
(450 B. C.)
4
Importance of Sensory Measurements

• Properties perceived by consumers of most
  importance to food quality
• Objective measurements calibrated against human
  standards
• Instruments not measure acceptability,
  likability, etc.
• Processes in the mouth may change nature of the
  food (saliva, temperature, enzymes)
• Mechanical actions of mouth difficult to duplicate

5
Disadvantages of Sensory Testing

• Humans vary widely in sensitivity, likes and
  dislikes
• Variables may be difficult to isolate (influence
  of color, taste, smell)
• May be expensive, time consuming
• Standards are not absolute

6
Types of Sensory Tests

• Overall difference tests
• Attribute difference tests
• Qualitative affective tests
• Quantitative affective tests
• Acceptance tests

7

• Sensory evaluation using people as measuring
  tools
• sensory evaluation panels
• identify criteria detect changes in product
• consumer panels
• assess product acceptability/desirability
• consumer feedback
• how does product match consumer demands

8

• Sensory evaluation panels
• highly trained
• judge quality of product using internal or
  industry standards
• untrained panels
• evaluate new products for acceptability
• evaluate effects of process or formula change
  (Are their differences?)

9
OVERALL DIFFERENCE TESTS

• Designed to detect whether subjects can detect
  any difference amongst samples
• Ex 1 Manufacturer substitutes carageenan for
  xanthan in yogurt. Can difference in texture be
  detected
• Ex 2 Juice producer testing non-thermal
  pasteurization. Perceived differences in
  texture?

10
Triangle Test

• Determine if overall differences between 2
  products
• Easy to do and analyze
• 20 - 40 panelists typical 8-12 for large
  differences
• Similarity tests 50 - 100 panelists

11
3 Samples are presented, two are the same, one is
different. Random combinations of ABB, BAA,
AAB, BBA, ABA, BAB are presented. Panelists are
asked to identify the odd sample. Subjects must
guess if they cannot detect a difference. The
number of correct responses are tabulated and
checked for significance.
A
B
A
12
Two Out of Five Test

• Also for overall differences
• Similar to triangle but more statistically
  efficient (smaller chance of guessing)
• Can use smaller number of panelists
• 5 coded samples must select two samples that are
  different

13
Duo-Trio Tests

• Simple, easy test, but less statitically
  efficient than triangle test
• May need gt 30 panelists
• Reference sample presented, followed by two coded
  samples
• Which sample matches reference?

14
(No Transcript)
15
Simple Difference Test

• Effective when only two samples can be presented
• Two samples presented
• Panelists indicate whether samples are the same
  or not
• Analysis compare the number of different
  responses for matched pairs with the number of
  different responses for different pairs
• Use ?2 for analysis

16
Difference From Control Test

• Determine difference between samples and a
  control
• Good when size of difference is important
• Panelists rate the size of the difference between
  each sample and the control

17
(No Transcript)
18
ATTRIBUTE DIFFERENCE TESTS

• Designed to detect whether a single attribute
  differes between, or amongst, samples
• Hardness, viscosity, crispness, etc

19
Paired Comparison Test

• Designed to determine in what way an attribute
  differs between samples (harder, less viscous,
  etc)
• Simple, useful for screening
• Often large number of panelists needed
• 2 coded samples presented which sample is
  harder, crisper, more cohesive, etc

20
Pairwise Ranking Test

• Used to compare several samples for a single
  attribute (3-6 samples by inexperienced
  panelists)
• Samples randomly presented, and in pairs

Which sample is crispier? A___ B___
21

• For a given pair, panelists asked which sample is
  more __________?
• Freidman ranked sum analysis used
• Results shown on a rank sum scale

D
C
A
B
Not Crisp
Very Crisp
40 50 60 70 80
22
Simple Ranking Tests

• Several samples compared at once for a single
  attribute
• Ranking data is ordinal
• Samples presented in random order

23
Panelists rank the sample according to the
attribute Rank sums are calculated and
evaluated by Friedmans test
Rank the samples in order from least to most
chewy. 1st__________
2nd_________ 3rd__________
4th__________
24
Rating/ANOVA Approach

• Panelists decide in what way an attribute varies
  over all the samples
• Rate attribut on a numerical scale
• Multiple samples presented in random order

25
Panelists rate attribute on a cardinal scale
Mark a line on the sliding scale for each sample
to indicate how hard they are. B C A
26
BIB Ranking Test

• Panelists determine in what way an attribute
  varies over all the samples
• Useful when many samples exist
• Samples presented in smaller groups according to
  a Cochran and Cox incomplete block design
• Samples ranked according to attribute

27
AFFECTIVE TESTS

• Used to determine consumers personal response to
  a product
• Preference does consumer like one product better
  than others
• Acceptance is the product one that a consumer
  finds acceptable, would buy, or would use
• Consumers may find 2 products acceptable, but not
  prefer one over another. Also, panelists might
  prefer one product over others, yet none of them
  may be acceptable.

28
Some uses . . .

• Product Maintenance Does ingredient, process, or
  packaging changes affect the characteristics and
  acceptability of existing product?
• Product Improvement Are attempts to improve a
  product through formulation or process effective?
• New Product Development Are new products being
  developed acceptable to consumers?
• Category Review How does an existing product
  stack up against competiotive brands?

29

• Conducted on group of subjects representative of
  larger target population
• Factors
• User group (e.g. sports drinks, baby foods)
• Age
• Gender
• Household income
• Geographic location
• Ethnicity, education, employment

30
Qualitative Affective Tests

• Measure subjective responses
• Assess consumers initial response
• Learn consumer terminology to describe products
  in their own words
• Learn how consumers use a product

31
Focus Groups

• Focus group consists of 10-12 consumers selected
  based on specific criteria
• Subjects meet with moderator 1-2 hrs
• Presents product and facilitates discussions
• Responses summarized in written form
• Often videotaped

32
Focus Panels

• Focus group in which the panel discusses the
  product, is sent home to use it, then returns for
  further discussion

33
One-On-One Interviews

• Consumers interviewed individually about their
  response to a product
• Similar format, set of questions, etc used for
  each interviewee

34
QUANTITATIVE AFFECTIVE TESTS

• Used to determine overall prefernce or liking for
  a product
• Often used to determine preference or liking for
  broad aspects of a product, but may be used to
  measure particular sensory attributes
• May be choice (preference) or rating (acceptance)

35
Preference Tests

• Useful when one product is pitted against another
• Improved product versus control
• Do not indicate degree of liking

36
NUMBER OF TEST SAMPLES
ACTION Paired Preference 2 Choose one
sample over Another (A/B) Rank Preference
3 or more Rank in order of preference (B-A
-C-D) Multiple Paired 3 or more Series of
paired samples. One Preference sample
chosen over the other for each
pairing. Multiple Paired 3 or more As above,
but series of samples Preference paired
with selected samples.
37
Acceptance Tests

• Determine how well a product is liked by
  consumers
• Often similar to the attribute difference tests
  discussed previously, but the attribute is
  acceptance or liking
• May be appropriate to infer prefence based on
  relative acceptance scores