Scribd
Upload
5 views

Measurement Response

Uploaded by

Xue Yi Lam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
5 views

Measurement Response

Uploaded by

Xue Yi Lam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 32

NT31103

Measuring
Response
DR NOORAKMAR AB WAHAB
[email protected]

Prepared by:
Dr. Ahmad Riduan Bahauddin
[email protected]
The real world of tasting and testing is not
simple, and a much more varied approach is
needed

Introduction
Tasters were not really measuring instruments;
we cannot set them up with a range of 0 to
100 and supply a couple of calibration points
for each attribute to be rated.
Measurement theory

Measurement theory These are referred to


tells us that numbers as nominal scaling,
can be assigned to ordinal scaling,
items in different interval scaling, and
ways. ratio scaling.
• In nominal scaling, numbers are assigned to events merely as labels,
contains the least information.

• The appropriate analysis of such data is to make frequency counts

• The only valid comparisons between individual items with this scale is to
say whether they belong to the same category or to different ones (an
equal versus not equal decision).
• In ordinal scaling, numbers are assigned to recognize the rank order of products with regard
to some sensory property, attitude, or opinion (such as preference).

• In this case increasing numbers assigned to the products represent increasing amounts or
intensities of sensory experience.

• In this case the numbers do not tell us anything about the relative differences among the
products.

• We cannot draw conclusions about the degree of difference perceived nor the ratio or
magnitude of difference.
• The next level of scaling occurs when the subjective spacing of responses is equal, so
the numbers represent equal degrees of difference.

• This is called interval-level measurement.

• Few scales used in sensory science have been subjected to tests that would help
establish whether they achieved an interval level of measurement and yet this level
is often assumed

• The advantage of interval-level measurement is that the data allow added


interpretation.
• A second advantage is that more powerful statistical methods may be
brought to bear the parametric methods. Computation of means, t-tests,
linear regression, and analysis of variance are appropriate analyses.

• Another even more desirable level of measurement is ratio


measurement. It has been widely assumed that the method of
magnitude estimation is a priori a ratio scaling procedure.
When we ask panelists to assign numbers or labels to sensory impressions, they may do this in
at least four ways :-

Nominal Ordinal Interval Ratio


• Items examined are • Items examined into • Items into numbered • use numbers which
placed in two or more two or more groups groups separated by a indicate how many
groups which belong to an constant interval times the stimulus in
• Do not obey any ordered series • can be analyzed by all questions stronger (or
particular order or any • e.g slight, moderate, nonparametric and saltier, or more
quantitative strong often by parametric irritating) than a
relationship • carry more methods reference
• Least Information information and can • Preferred by because
be analyzed by most they are free from
nonparametric end-of-scale
statistical tests distortions
Classification
→ Sorted into group
Most frequently
used methods
Ranking
of measuring
→Arranged in order of intensity or degree
sensory
response to a
Grading
sample are, in
→depend on expert graders
order of
increasing
Scaling
complexity:
→ judge the sample by reference to a scale of
numbers
a. Classification

• Subjects are asked to select an attribute or attributes which describe the stimulus

• No attempt is made to standardize the terms, and the results are reported as the
number of check marks for each term
• Such data are nominal: no numbers are used, and there is no increasing or decreasing
series expressed in the data
b. Grading

• Method of evaluation much used in commerce which


depends on expert “graders”
• Examples of items subjected to sensory grading are coffee,
tea, spices, butter, fish, and meat
• Useful in commerce where they protect the consumer
against being offered low-quality products at a high price
• Suffers from the considerable drawback that statistical
correlation with measurable physical or chemical
properties is difficult or impossible
c. Ranking

• Subjects receive three* or more samples which are to be


arranged in order of intensity or degree of some specified
attribute
• For example, four samples of yogurt are to be ranked for
degree of sensory acidity, or five samples of breakfast
cereal may be ranked for preference
• Rapid and demand relatively little training but subjects
must be thoroughly familiarized with the attribute under
test
d. Scaling

• Application of numbers, or judgments that are


converted to numerical values,
• Describe the perceived intensity of a sensory
experience or the degree of liking or disliking for some
experience or product.
• A variety of scale have been developed for this purpose
and with some caution, all work well in differentiating
products.
• The validity and reliability of a scaling technique are highly dependent upon:-
✓ Broad enough to encompass the full range of parameter intensities and also has
enough discrete points to pick up all the small differences in intensity between
samples
✓ The degree to which the panel has or has not been taught to associate a
particular sensation (and none other) with the attribute being scaled
✓ The degree to which the panel has or has not been trained to use the scale in
the same way across all samples and across time
Category Scales

• The oldest method of scaling involves the choice of discrete response


alternatives to signify increasing sensation intensity or degrees of liking
and/or preference.

• A method of measurement in which the subject is asked to “rate” the


intensity of a particular stimulus by assigning it a value (category) on a
limited - generally considered to be at least ordinal level data
• In a category scale the number of alternative responses is limited - do not generally
provide values which measure the degree (how much) one sample is more than another.

• Panelists may also tend to use the categories with equal frequency and usually avoid the
use of the two scale end points in order to save them for “real extremes.”

• A key idea is to present an easily understandable word like “sweetness” and ask the
participant to evaluate the perceived intensity of that attribute.
• Disadvantages: -
➢Prone to have ‘ceiling effect’.
➢Response cannot be calibrate among the panellists
Pictorial Scales

• Preference or acceptance testing with children can be done with a few modifications
from the adult methods.
• These often include the following:
1) one on-one testing in most cases, to insure compliance, understanding, and to
minimize social influences,
2) children can respond to either verbal scales or pictorial scales,
3) scales may need to be truncated for use with younger children,
4) paired preference testing is suitable for very young children in the ranges about
4–5 years.
Line Scale

• A second widely used technique for intensity scaling involves making a mark or
slash on a line to indicate the intensity of some attribute.

• The response is recorded as the distance of the mark from one end of the scale,
usually whatever end is considered “lower.”

• Line marking differs from category scales in the sense that the person’s choices
seem more continuous and less limited.
• The fundamental idea is that the panelist makes a mark on a line to indicate the intensity
or amount of some sensory characteristic.

• Interval scale consisting of a horizontal line 6 in. or 15 cm long with anchor points 0.5 in.
(1.3 cm) from each end and usually but not necessarily having a mid point.

• 2 type of line scale :-


a. Structured scale – scale with ↑/↓ intensity label along the scale
b. Unstructured scale – wide scale without any label except the anchor point
• Line scales are sometimes referred to as visual analog scales (or VAS).

• Advantage is that the intensity can be more accurately graded because there are
no steps or favourite numbers; disadvantage - it is harder for a panelist to be
consistent.

• Recommend the use of line marking for Quantitative Descriptive Analysis (QDA)
Magnitude Estimation

• In this procedure, the respondent is instructed to assign numbers to sensations


in proportion to how strong the sensation seems.

• Specifically, the ratios between the numbers are supposed to reflect the ratios
of sensation magnitudes that have been experienced.

• For example, if product A is given the value of 20 for sweetness intensity and
product B seems twice as sweet, B is given a magnitude estimate of 40.
The two critical parts of the technique are the instructions given to the participant and the techniques for
data analysis.

Two primary variations of magnitude estimation have been used. In one method, a standard stimulus is
given to the subject as a reference and that standard is assigned a fixed value such as 10.

In the other variation of magnitude estimation, no standard stimulus is given and the participant is free
to choose any number he or she wishes for the first sample.

i. Panellists should have basic mathematical skill


Disadvantages :-
ii. Thorough explanation on scale usage need to be provided
Other scales
 Kemp, S. E., Hollowood, T. & Hort, J. 2009. Sensory
Evaluation: A Practical Handbook. United Kingdom: Wiley-
Blackwell

 Lawless, H.T. & Heymann, H. 2010. Sensory Evaluation of


Food: Principles and Practices. 2nd Ed. New York:
Springer.
References
 Meilgaard, M., Civille, G.V. & Carr, B.T. 2007. Sensory
Evaluation Techniques. 4th Ed. Boca Raton, Florida: CRC
Press.

You might also like