Uploaded by Dhruv Joshi

measurement and scaling

advertisement
SESSION
Measurement and Scaling:
Fundamentals and
Comparative Scaling
SESSION: Outline
•
Measurement and Scaling
•
Primary Scales of Measurement
•
A Comparison of Scaling
Techniques
•
Comparative Scaling Techniques
Measurement and Scaling
Measurement means assigning numbers or other symbols to
characteristics of consumers according to certain pre-specified
rules.
 Measure consumer perceptions, attitudes, preferences, and
other relevant characteristics.
 The rules for assigning numbers should be standardized and
applied uniformly.
• Measurement is assigning range/ points across which
respondents can not vary in their response
Measurement and Scaling
cont...
• Scaling is an extension of measurement.
• Scaling involves plotting consumers according to
their responses
e.g. Consider an attitude scale from 1 to 5. Each
respondent is assigned a number from 1 to 5
Where: 1 = Extremely Unfavorable
5 = Extremely Favorable
• Measurement is the actual assigning range from 1 to
5 to each respondent and Scaling is the placing of
respondents on these numbers with respect to their
attitude.
Primary Scales of Measurement
Nominal Scale
• The numbers serve only as labels or tags for identifying and
classifying objects.
• When used for identification, there is a strict one-to-one
correspondence between the numbers and the objects.
• The numbers do not reflect the amount of the
characteristic possessed by the objects.
• The only permissible operation on the numbers in a nominal
scale is counting.
• Only a limited number of statistics, all of
which are based on frequency counts, are
permissible, e.g. percentages and mode.
Primary Scales of Measurement
Ordinal Scale
• A ranking scale in which numbers are assigned to objects to indicate the relative
extent to which the objects possess some characteristic and also categories of
objects.
• Can determine whether an object has more or less of a characteristic than some
other object, but not how much more or less.
• In addition to the counting operation allowable for nominal scale data,
ordinal scales permit the use of statistics based on centiles, e.g.,
percentile, quartile, median.
Primary Scales of Measurement
Interval Scale
• Numerically equal distances on the scale represent equal
values in the characteristic being measured.
• In Interval scale there is no absolute Zero scale point.
• It permits comparison of the differences between objects.
• Magnitude of differences can be captured and can
comparing the differences in attributes.
• Statistical techniques that may be used include all of those that
can be applied to nominal and ordinal data, and in addition
the arithmetic mean, standard deviation, and other statistics
commonly used in marketing research.
Nominal, Ordinal and Interval Scale
• Country E refers to Croatia,
assigned to denote the country
and serve the purpose of
identification and are not
related to their football
playing capabilities.
- Nominal
• Croatia, ranked 5, played
better than Turkey, ranked 10.
- Ordinal
• The magnitude of the
difference can be obtained
only by looking at the points
i.e. difference between 1,266
points and 1,033 points
respectively. - Interval
Nominal
Team
ID
Ordinal
Interval
Rank
Points
A
Spain
1
1,565
B
Italy
2
1,339
C
Germany
3
1,329
D
Netherlands 4
1,295
E
Croatia
5
1,266
F
Brazil
6
1,252
G
Argentina
7
1,230
H
Czech Rep.
8
1,134
I
Portugal
9
1,120
J
Turkey
10
1,033
Primary Scales of Measurement
Ratio Scale
• Possesses all the properties of the nominal, ordinal, and interval
scales:
– Can identify or classify objects (nominal nature)
– Rank the objects (ordinal nature)
– Compare intervals or differences (interval nature)
• It has an absolute zero point. Absolute zero represents a point on the
scale where there is an absence of the given attribute.
• All statistical techniques can be applied to ratio data.
• Ratio scales determines the absolute rather than the relative
strengths of the attribute.
Illustration of Primary Scales of Measurement
Nominal
Scale
Ordinal
Scale
Interval
Scale
Ratio
Scale
No. Store
Preference
Rankings
Preference
Ratings
1-7
$ spent last
3 months
1. Parisian
2. Macy’s
3. Kmart
4. Kohl’s
5. J.C. Penney
6. Neiman Marcus
7. Marshalls
8. Saks Fifth Avenue
9. Sears
10.Wal-Mart
7
2
8
3
1
5
9
6
4
10
5
7
4
6
7
5
4
5
6
2
0
200
0
100
250
35
0
100
0
10
Primary Scales of Measurement
Scale
Nominal
Ordinal
Interval
Ratio
Basic
Characteristics
Numbers identify
& classify objects
Common
Examples
Social Security
nos., numbering
of football players
Nos. indicate the Quality rankings,
relative positions rankings of teams
of objects but not in a tournament
the magnitude of
differences
between them
Differences
Temperature
between objects
(Fahrenheit)
Zero point is fixed, Length, weight
ratios of scale
values can be
compared
Marketing
Examples
Brand numbers,
store types
Preference
rankings, market
position, social
class
Attitudes,
opinions, index
sales, income, &
costs
Primary Scales of Measurement (Source: Churchill & Brown, Basic MR, 6e, ISE, 2007
A Classification of Scaling Techniques
Scaling Techniques
Noncomparative
Scales
Comparative
Scales
Paired
Comparison
Rank
Order
Continuous
Itemized
Rating Scales Rating Scales
Constant
Sum
Likert
Semantic
Differential
Stapel
A Comparison of Scaling Techniques
• Comparative scales involve the direct comparison
between options. e.g. What you prefer b/w Pepsi and
Coca-Cola.
Comparative scale data must be interpreted in
relative terms and have only ordinal or rank order
properties.
• In noncomparative scales, each object is scaled
independently of the others in the option set. The
resulting data are generally assumed to be interval
or ratio scaled.
Comparative Scaling Techniques
Paired Comparison Scaling
• A respondent is presented with two objects and
asked to select one according to some criterion.
• The data obtained are ordinal in nature.
• Paired comparison scaling is the most widely used
comparative scaling technique.
• Paired Comparison data can be converted to a Rank
order data.
Obtaining Shampoo Preferences
Using Paired Comparisons
Instructions: We are going to present you with ten pairs of
shampoo brands. For each pair, please indicate which one of the two
brands of shampoo you would prefer for personal use.
Head &
Pert
Recording Form: Jhirmack Finesse Vidal
Jhirmack
aA
0
Sassoon Shoulders
0
1
Finesse
1a
Vidal Sassoon
1
1
Head & Shoulders
0
0
0
Pert
1
1
0
1
Number of Times
Preferredb
3
2
0
4
0
1
1
0
0
1
0
1
1 in a particular box means that the brand in that column was preferred
over the brand in the corresponding row. A 0 means that the row brand was
preferred over the column brand. bThe number of times a brand was preferred
is obtained by summing the 1s in each column.
Paired Comparison Selling
The most common method of taste testing is paired comparison. The consumer is
asked to sample two different products and select the one with the most appealing
taste. The test is done in private and a minimum of 1,000 responses is considered
an adequate sample. A blind taste test identifies the imagery, self-perception
and brand reputation in the consumer’s mind to predict the probable purchasing
decision.
A paired comparison
taste test
Comparative Scaling Techniques
Rank Order Scaling
• Respondents are presented with several
objects simultaneously and asked to order or
rank them according to some criterion.
• Measure preferences for brands and attributes.
• Rank order scaling forces the respondent to
discriminate among various brands.
Preference for Toothpaste Brands
Using Rank Order Scaling
Instructions: Rank the various brands of toothpaste in order of
preference. Begin by picking out the one brand that you like most
and assign it a number 1. Then find the second most preferred
brand and assign it a number 2. Continue this procedure until you
have ranked all the brands of toothpaste in order of preference.
The least preferred brand should be assigned a rank of 10.
No two brands should receive the same rank number.
Preference for Toothpaste Brands
Using Rank Order Scaling
Form
Brand
Rank Order
1. Crest
_________
2. Colgate
_________
3. Aim
_________
4. Gleem
_________
5. Macleans
_________
6. Ultra Brite
_________
7. Close Up
_________
8. Pepsodent
_________
9. Plus White
_________
10. Stripe
_________
Comparative Scaling Techniques
Constant Sum Scaling
• Respondents allocate a constant sum of units, such
as 100 points to attributes of a product to reflect
their importance.
• If an attribute is unimportant, the respondent assigns
it zero points.
• If an attribute is twice as important as some other
attribute, it receives twice as many points.
• The sum of all the points is 100. Hence, the name of
the scale.
Importance of Bathing Soap Attributes
Using a Constant Sum Scale
Instructions
On the next slide, there are eight attributes of bathing
soaps. Please allocate 100 points among the attributes
so that your allocation reflects the relative importance
you attach to each attribute. The more points an
attribute receives, the more important the attribute is.
If an attribute is not at all important, assign it zero
points. If an attribute is twice as important as some
other attribute, it should receive twice as many points.
Importance of Bathing Soap Attributes
Using a Constant Sum Scale
Form
Average Responses of Three Segments
Attribute
1. Mildness
2. Lather
3. Shrinkage
4. Price
5. Fragrance
6. Packaging
7. Moisturizing
8. Cleaning Power
Sum
Segment I
8
2
3
53
9
7
5
13
100
Segment II
2
4
9
17
0
5
3
60
100
Segment III
4
17
7
9
19
9
20
15
100
SESSION
Measurement and Scaling:
Non comparative Scaling
Techniques
SESSION: Outline
•
•
Non comparative Scaling Techniques
•
Continuous Rating Scale
•
Itemized Rating Scale
Non comparative Itemized Rating Scale
Decisions
Noncomparative Scaling Techniques
• Respondents evaluate only one object at a
time.
• Noncomparative techniques consist of:
– Continuous and
– Itemized rating scales.
Continuous Rating Scale
Respondents rate the objects by placing a mark at the appropriate
position on a line that runs from one extreme of the criterion variable
to the other.
The form of the continuous scale may vary considerably.
How would you rate Sears as a department store?
Version 1
Probably the worst - - - - - - -I - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Probably the best
Version 2
Probably the worst - - - - - - -I - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -- - Probably the best
0 10
20
30
40
50
60
70
80
90
100
Version 3
Very bad
Neither good
Very good
nor bad
Probably the worst - - - - - - -I - - - - - - - - - - - - - - - - - - - - - -- - - - - - - - - - - - - - - - -Probably the best
0 10
20
30
40
50
60
70
80
90
100
RATE: Rapid Analysis and Testing Environment
A relatively new research tool, the perception analyzer, provides continuous
measurement of “gut reaction.” A group of up to 400 respondents are presented
with TV or radio spots or advertising copy. The measuring device consists of a
dial that contains a 100-point range. Each participant is given a dial and
instructed to continuously record his or her reaction to the material being tested.
As the respondents turn the dials, the
information is fed to a computer, which tabulates
second-by-second response profiles. As the
results are recorded by the computer, they are
superimposed on a video screen, enabling the
researcher to view the respondents' scores
immediately. The responses are also stored in a
permanent data file for use in further analysis.
The response scores can be broken down by
categories, such as age, income, sex, or product
usage.
Itemized Rating Scales
• The respondents are provided with a scale that
has a number or brief description associated
with each category.
• The categories are ordered in terms of scale
position, and the respondents are required to
select the specified category that best
describes the object being rated.
• The commonly used itemized rating scales are
the Likert, semantic differential, and Stapel
scales.
Likert Scale
The Likert scale requires the respondents to indicate a degree of
agreement or disagreement with each of a series of statements about
the stimulus objects.
Strongly
disagree
Disagree
Neither
agree nor
disagree
Agree
Strongly
agree
1. Sears sells high quality merchandise.
1
2X
3
4
5
2. Sears has poor in-store service.
1
2X
3
4
5
3. I like to shop at Sears.
1
2
3X
4
5
• The analysis can be conducted on an item-by-item basis (profile
analysis), or a total (summated) score can be calculated.
Semantic Differential Scale
•
•
•
•
•
•
•
The semantic differential is a seven-point rating scale with end points associated with bipolar
labels that have semantic meaning.
SEARS IS:
Powerful
--:--:--:--:-X-:--:--: Weak
Unreliable
--:--:--:--:--:-X-:--: Reliable
Modern
--:--:--:--:--:--:-X-: Old-fashioned
The negative adjective or phrase sometimes appears at the left side of the scale and sometimes at
the right.
This controls the tendency of some respondents, particularly those with very positive or very
negative attitudes, to mark the right- or left-hand sides without reading the labels.
Individual items on a semantic differential scale may be scored on either a -3 to +3 or a 1 to 7
scale.
Mark the blank that best indicates how accurately one or more words/ adjectives
describes the characteristics of an object (i.e. aspect to be studied – co., brand etc.)
Profile Analysis is used here – each profile/bipolar words are separately analyzed. Means on
each rating scale are calculated and compared by plotting the graph.
Determines the overall differences and/or similarities among the objects/ respondents.
Assess difference across segment of respondents – compare mean responses of different
segments.
A Semantic Differential Scale for Measuring
Self- Concepts, Person Concepts, and Product Concepts
1) Rugged
:---:---:---:---:---:---:---: Delicate
2) Excitable
:---:---:---:---:---:---:---: Calm
3) Uncomfortable
:---:---:---:---:---:---:---: Comfortable
4) Dominating
:---:---:---:---:---:---:---: Submissive
5) Thrifty
:---:---:---:---:---:---:---: Indulgent
6) Pleasant
:---:---:---:---:---:---:---: Unpleasant
7) Contemporary
:---:---:---:---:---:---:---: Obsolete
8) Organized
:---:---:---:---:---:---:---: Unorganized
9) Rational
:---:---:---:---:---:---:---: Emotional
10) Youthful
:---:---:---:---:---:---:---: Mature
11) Formal
:---:---:---:---:---:---:---: Informal
12) Orthodox
:---:---:---:---:---:---:---: Liberal
13) Complex
:---:---:---:---:---:---:---: Simple
14) Colorless
:---:---:---:---:---:---:---: Colorful
15) Modest
:---:---:---:---:---:---:---: Vain
Stapel Scale
The Stapel scale is a unipolar rating scale with ten categories numbered from -5
to +5, without a neutral point (zero). This scale is usually presented vertically.
SEARS
+5
+4
+3
+2
+1
HIGH QUALITY
-1
-2
-3
-4X
-5
+5
+4
+3
+2X
+1
POOR SERVICE
-1
-2
-3
-4
-5
• Evaluate how accurately each word/ phrase describes the
characteristics.
Basic Noncomparative Scales
Scale
Basic
Characteristics
Examples
Advantages
Disadvantages
Place a mark on a
continuous line
Reaction to
TV
commercials
Easy to construct
Scoring can be
cumbersome
unless
computerized
Degrees of
agreement on a 1
(strongly disagree)
to 5 (strongly agree)
scale
Measurement
of attitudes
Easy to construct,
administer, and
understand
More
time - consuming
Semantic
Differential
Seven - point scale
with bipolar labels
Brand,
product, and
company
images
Versatile
Controversy as
to whether the
data are interval
Stapel
Scale
Unipolar ten
- point
scale, - 5 to +5,
without a neutral
point (zero)
Measurement
of attitudes
and images
Easy to construct,
administer over
telephone
Confusing and
difficult to apply
Continuous
Rating
Scale
Itemized
Rating
Scales
Likert Scale
(As respondents have to
read statements in detail)
Some Commonly Used Scales in Marketing
CONSTRUCT
SCALE DESCRIPTORS
Attitude
Very Bad
Bad
Neither bad
nor Good
Good
Very Good
Importance
Not at All
important
Not
important
Neutral
Important
Very
Important
Satisfaction
Very
Dissatisfied
Dissatisfied
Neither
dissatisfied
nor satisfied
Satisfied
Very Satisfied
Purchase
Intent
Definitely
Probably
will Not Buy will Not
Buy
Might or
Might Not
Buy
Probably
Will Buy
Definitely
Will Buy
Purchase
Frequency
Never
Sometimes
Often
Very Often
Rarely
Non-comparative
Itemized Rating Scale decisions
Six major decisions when constructing noncomparative itemized rating scales:
• Number of scale categories to use
• Balanced vs. Unbalanced scale
• Odd or Even number of categories
• Forced vs. Non-forced choice
• Nature and Degree of Verbal description
• Physical form of the Scale
Number of scale categories
• Trade-offs required: greater the number of scale categories, finer the
discrimination among respondents/ objects; whereas respondents have a
limit of handling categories.
• Traditional guidelines:
– 7±2 = 5 or 9 categories are ideal.
– If the respondent is knowledgeable about the study, a larger scale categories give
them better choice for discrimination, whereas if respondent are not much aware
then fewer categories should be used to avoid confusion.
– Nature of objects: in some objects finer discrimination is difficult – small scale
categories should be used e.g. attitude or choice behavioral studies.
– Mode of data collection: if done through telephonic survey – lesser categories are
suitable whereas use maximum categories in personal surveys.
– Type of data analysis: if higher level of data analysis is needed, 7 or more
categories are recommended (As size of the Correlation coefficient – measure of
relationship – is influenced by the number of scale categories – Coefficient
decreases with reduction in scale categories - further impact on other higher
level analysis).
Balanced vs. Unbalanced Scales
• Balanced scale:
number of favourable
and unfavourable
categories are equal.
Balanced Scale
Unbalanced Scale
Surfing the Internet is
Surfing the Internet is
____ Extremely Good
____ Extremely Good
____ Very Good
____ Very Good
____ Good
____ Good
____ Bad
____ Somewhat Good
____ Very Bad
____ Bad
____ Extremely Bad
____ Very Bad
Odd vs. Even number of Scales &
Forced vs. Non forced scales
• Odd number of categories: middle scale position is
designated as neutral/ impartial.
e.g. Likert scale is the most balanced rating scale
proposed with odd number of categories (5) and a
neutral point.
• Odd categories are useful when possibility of neutral
responses is there.
• When respondent is forced to respond on categories
and no response option is not there – even number
of scale categories are used.
Nature and Degree of Verbal Description
• Scale categories may have verbal, numerical or
pictorial descriptions.
• It is recommended – to have verbal description for
each category to reduce scale ambiguity.
Physical Form or Configuration
Rating Scale Configurations
Cheer detergent is:
1) Very harsh
---
2) Very harsh
1
--2
---
---
---
---
---
Very gentle
3
4
5
6
7
Very gentle
3) . Very harsh
.
Cheer
.
. Neither harsh nor gentle
.
.
. Very gentle
4) ____
____
____
____
Very
Harsh
Somewhat Neither harsh
harsh
Harsh
nor gentle
5) -3
Very
harsh
-2
-1
0
Neither harsh
nor gentle
____
Somewhat
gentle
+1
____
Gentle
____
Very
gentle
+2
+3
Very
gentle
Some Unique Rating Scale Configurations
Thermometer Scale
Instructions: Please indicate how much you like McDonald’s hamburgers by coloring in
the thermometer. Start at the bottom and color up to the temperature level that best
indicates how strong your preference is.
Form:
Like very
much
Dislike
very much
100
75
50
25
0
Smiling Face Scale
Instructions: Please point to the face that shows how much you like the Barbie Doll. If
you do not like the Barbie Doll at all, you would point to Face 1. If you liked it very much,
you would point to Face 5.
Form:
1
2
3
4
5
Measurement Accuracy
•
Measurement is not the complete representation of the characteristics of studied object but
an observation of it. So, it may go wrong. This causes Measurement error.
•
This results in the measurement or observed score being different from the true score of the
characteristic (e.g. behaviour) being measured. The True Score Model is a mathematical model
that provides the framework for understanding the accuracy of measurement.
X O X T  X S  X R 
where
XO = the observed score or measurement
XT = the true score of the characteristic
XS = systematic error
XR = random error
Measurement Accuracy
cont…
Two possible errors may occur:
• Systematic/Constant errors: affects the
measurement always in a constant way. It represents
the stable factors that affect the observed score in the
same way every time.
• Random errors: factors that affect the observed score
in different ways each time based on situations. e.g.
noise/ distractions, emotions, fatigue etc.
Potential Sources of Error in Measurement
• Stable characteristics of the individual that influence the test score mostly:
intelligence, social background and education.
• Short-term or transient personal factors: health, emotions and fatigue.
• Situational factors: presence of other people, noise, and distractions.
• Sampling of items included in the scale: addition, deletion, or changes in the scale
items.
• Lack of clarity of the scale: inadequate instructions on the items.
• Mechanical factors: poor printing, overcrowding items in the questionnaire, and poor
design.
• Administration of the scale: differences among interviewers.
• Analysis factors: fault in scoring and using the statistical analysis.
Measurement Accuracy cont…
• In the long run, the expected mean of measurement errors
should be zero. When the error term is zero, the observed
score is the true score:
XO = XT + 0;
XO = XT
• Therefore, if an instrument is reliable, the observed scores
should be fairly consistent and stable throughout
repeated measures. Many popular reliability estimate
methods such as Cronbach’s coefficient alpha, test-retest
are based upon this rationale.
Reliability V/S Validity
• Reliability means the extent to which scaling results are free from
experimental (random) errors i.e. XR => extent to which a scale
produces consistent scores/ results, if measurements are repeated.
(factor of ‘trust’)
• Systematic errors do not have adverse impact on reliability – as they
effect measurement in a constant way and thus do not lead to
inconsistency.
• Random errors produces inconsistency – leads to lower reliability.
• Validity means that the data obtained must be unbiased and relevant
to describe the characteristics being measured – extent to which
differences in observed scale scores reflect true differences among
objects/ respondents response on the characteristics being
observed. [ XO = XT ]
• Reliability is a necessary, but not sufficient, condition for validity –
because Systematic error may still be there. [ XO = XT + XS ]
Reliability V/S Validity
• Reliability refers to the repeatability of findings. If the study
were to be done a second time, would it yield the same
results? If so, the data are reliable.
• If more than one person is observing behavior or some
event, all observers should agree on what is being recorded
in order to claim that the data are reliable. E.g. When
people take a vocabulary test two times, their scores on
the two occasions should be very similar.
Reliability V/S Validity
• Validity refers to the credibility or believability of the
research. Are the findings genuine?
• There are two aspects of validity:
• Internal validity - the instruments or procedures used in
the research measured what they were supposed to
measure. Example: As part of a stress experiment, people
are shown photos of war killings. After the study, they are
asked how the pictures made them feel, and they respond
that the pictures were very upsetting. In this study, the
photos have good internal validity as stress producers.
• External validity - the results can be generalized beyond
the immediate study. It should also apply to people beyond
the sample in the study.
Measurement Accuracy
Scale Evaluation
Validity
Reliability
Test/
Retest
Alternative
Forms
Internal
Consistency
Content
Criterion
convergent
Construct
discriminant
Test – Retest Reliability
• Suppose, we measure Job Satisfaction among sales
executives at the first week and then measure it
again two weeks later.
Condition:
• If measure of job satisfaction is ‘reliable’, the two
scores must be highly correlated.
• If not, the measure is ‘unstable’ and unreliable.
• This is Known as test – retest reliability.
Limitations of Test – Retest Reliability
1. It is sensitive to the time interval of two
measurements i.e. longer the time interval, lower
the reliability.
2. The initial measurement may alter the
characteristic being measured i.e. measuring
attitude towards low-fat milk may cause them to
become more health conscious and their later
response may be biased due to the positive attitude
towards low fat milk.
3. Respondent may attempt to remember
responses and may give a biased/ polished
response next time.
Alternative – Forms Reliability
• Two equivalent form of scales are constructed.
• Same respondent is measured at two different time intervals (usually
2-4 weeks).
• Scores are correlated to assess the reliability.
E.g. Highly satisfied – Slightly satisfied – Neither satisfied nor
dissatisfied – Slightly Dissatisfied – Highly Dissatisfied
and
Very Satisfied – Satisfied – Neither satisfied nor dissatisfied –
Dissatisfied – Very Dissatisfied
Limitation:
• It is time consuming and expensive to construct equivalent type of
scales i.e. equivalent w.r.t. the content in the scale and they should
have the same meaning, variance and intercorrelations.
Internal Consistency Reliability
• Internal consistency reliability is a measure of how well a test
addresses different constructs and delivers reliable scores.
For example, Store image is a multidimensional construct that
includes: quality of merchandise, variety, assortments, service, price
offers, convenience of location, store layout, and credit and billing
facilities.
• The internal consistency reliability test provides a measure that
each of these particular constructs are measured correctly and
reliably. measure of how well a test addresses different constructs and
delivers reliable scores.
Measure 1: SPLIT-HALVES TEST
• For example, a questionnaire to measure attitude could be divided into
odd and even questions. The results from both halves are statistically
analysed, and if there is weak correlation between the two, then there
is a reliability problem with the test.
The division of the question into two sets must be random.
Internal Consistency Reliability
• Each item measures some aspect of the belief
construct i.e. 1-5 (strongly agree & strongly
disagree are among five items in 5 point scale).
• Items used in the scale should be consistent to
measure all aspects of characteristics i.e.
behaviour/attitude etc.
Measure 2:
Cronbach’s Alpha
Cronbach's
alpha
Internal
consistency
α ≥ .9
Excellent
• Coefficient alpha/ Cronbach’s Alpha is used to
measure this reliability and widely used.
.9 > α ≥ .8
Good
.8 > α ≥ .7
Acceptable
• This coefficient varies b/w 0 to 1 and value of ≤
0.6 indicates unsatisfactory internal consistency
reliability.
.7 > α ≥ .6
Questionable
.6 > α ≥ .5
Poor
.5 > α
Unacceptable
• Thus, this measure focuses on internal consistency
of scale points in the scale type.
For example: a series of questions might ask the respondent to rate their response between one
and five. Cronbach's Alpha gives a score of between zero and one, with 0.7 generally accepted as
a sign of acceptable reliability.
The test also takes into account both the size of the sample and the number of potential responses.
A 40-question test with possible ratings of 1 – 5 is seen as having more accuracy than a tenquestion test with three possible levels of response.
Validity
Perfect validity requires that there be no measurement error.
• Content or Face validity is a subjective but systematic evaluation of
how well the content of a scale represents the measurement and
whether the scale items adequately cover the entire domain of the
construct being measured. E.g. a scale designed to measure store
image would be inadequate, if it omitted any major dimension like
quality, variety or cleanliness etc.
• Criterion validity reflects whether a scale performs as expected in
relation to the variables selected (criterion variables). It may include
demographic and psychographic characteristics; attitudinal and
behavioural measures. Based on the time period involved, it may
take two forms:
– Concurrent validity
– Predictive validity
Criterion Validity
• Predictive Validity: Collects data on the scale at
one time and data on criterion variables at a future
time. E.g. attitude and intensions towards a
commodity brand captured by survey can be used to
predict future purchases of a scanner panel. Then
their future purchases are tracked by scanner data
(observation). The predicted and the actual
purchases are compared to assess the predictive
validity of attitudinal/ intension scale.
• Concurrent Validity: this validity is assessed when
data on the scale is evaluated and at the same time
data on the criterion variables are collected and then
both are compared.
Construct Validity
• Most difficult type of Validity to establish.
• Theoretical model is constructed based on the variables to
be measured.
• This been tested out with further research to either reject or
accept or modify the model.
• Construct validity defines how a well a test or experiment
measures up to its claims.
• Construct validity is a device used almost exclusively in social
sciences, psychology and education. E.g. research might
design to measure whether an educational program increases
artistic ability amongst pre-school children.
Construct Validity
Example:
• Construct validity is a measure of how well the test
measures the construct.
• Construct validity is valuable in social sciences, where there is
a lot of subjectivity to concepts. Often, there is no accepted
unit of measurement for constructs and even fairly well
known ones, such as IQ, are open to debate.
CONVERGENT AND
DISCRIMINANT VALIDITY
• The basic difference between convergent and discriminant
validity is that Convergent validity tests whether constructs
that should be related, are related. Discriminant validity tests
whether believed unrelated constructs are, unrelated.
E.g. Imagine that a researcher wants to measure self-esteem,
but also knows that the other
four constructs are related to
self-esteem, and have some
overlap. The ultimate goal is
to attempt to isolate
self-esteem.
CONVERGENT AND
DISCRIMINANT VALIDITY
• Convergent validity would test that the four other
constructs are, related to self-esteem in the study.
• Discriminant validity would ensure that, in the
study, the non-overlapping factors do not overlap.
For example, self esteem and intelligence (not in the
figure) should not relate.
HOW TO MEASURE CONSTRUCT VARIABILITY?
• For major and extensive research, especially in education and language
studies, most researchers test the construct validity before the main
research.
• These Pilot Studies establish the strength of their research and allow
them to make any adjustments.
• The other option is an intervention study, where a group with low
scores in the construct is tested, taught the construct, and then remeasured. E.g. testing the knowledge level of a subject to the students.
– If there is a significant difference pre and post-test, usually
analyzed with simple statistical tests, then this proves good
construct validity.
Download