Measurement and Scaling in Marketing Research

1
Measurement and Scalingin Marketing Research

• Chapter 10

2
Definition

• Scaling is the generation of a broadly defined
  continuum on which measured objects are located.

http//www.socialresearchmethods.net/kb/scalgen.ht
m
3
How Do We Collect Data for Advanced Analyses?

• Ordinal
• Paired comparisons
• Ranking procedures
• Ordered-category sorting
• Rating techniques
• Ratio/Interval
• Direct Judgment
• Fractionation
• Constant sum

4
Ordinal Paired Comparisons

• The judgments may not be transitive (circular
  triads).
• The judgments may not be consistent (A is
  preferred to B on one trial and B to A on
  another)
• The proportion of times that X appeared greater
  than Y

5
Ordinal

• Ranking Procedures
• From the following list please choose the three
  most important factors which attracted you to
  shop at
• Order k out of n.
• Ordered-Category Sorting
• Sort 15 detergents into three sudiness
  categories (1) high suds, (2) moderate suds,
  (3) low suds.
• Assumption of equal intervals separating
  boundaries.

6
Ordinal Rating Techniques

• Flexibility of rating - responses can be treated
  as ordinal as well as interval/ratio.
• Monadic (each object is measured by itself,
  independently of the other objects rated) or
  comparative (do you prefer A or B).
• Risk of little differentiation among scores. (end
  piling)
• Forms numerical, graphic and verbal.
• Examples gentleness/harshness, feelings
  thermometer, statement check.

7
Ordinal Scale Development

• Define the Construct
• Generate Items
• Test the Items
• Construct the Scale
• The researcher assumes that the descriptive
  levels of progress are in equal-interval steps
  psychologically.
• Other issues
• Number of categories
• Neutral scale item
• Balanced /unbalanced scale
• Descriptive adjectives,
• Anchoring phrases

8
Ratio/Interval Scales

• Direct Judgment
• Rate the taste of the food at TGI Fridays from
  1 (I dont like at all) to 7 (I like very much)
• Fractionation
• Compare each restaurant on the list to TGI
  Fridays, where TGI Fridays is 1.0
• Constant sum
• Assign 100 points across the five restaurants
  so as to reflect your relative degree of liking
  them.

9
Techniques for Scaling Stimuli

• 1) Thurstone Case V Scaling popular model used
  in dealing with ordinal data obtained from
  ranking methods.
• Uses Paired Comparisons Data (Do you prefer
  Brand a or b? All possible pairs given)
• Converts ranks into scale values

Initial Data Matrix Brand Preference Percent
6 7 5 4 3
2 1 6 .5000 .8900 .7200
.7200 .8300 .9000 .8000 7 .2200 .5000
1.0000 .6100 .9400 1.0000 .9000 5 .7200
.7800 .5000 .9400 .6700 .5000 .7000 4
.6700 .8900 .5000 .5000 .8300 .8000
.7000 3 .6700 .4400 .5600 .6700 .5000
.8000 .8000 2 .5000 .5000 .8300 1.0000
.5600 .5000 .8000 1 .7800 .5600 .7200
.5000 .6100 .7000 .5000
Resulting ScaleOverall Cat Food Preference 7
Brands .17 Prance Base Chicken

.16
Prance 9 mo (old)

.13 Prance 3 mo
(fresh) Whiskers Soft

.12 Kitty Lickins Soft

.08 New Product
V. 2

.00 New Product V.1


FINAL DRAWING ON 1.00 SCALE
10
Techniques for Scaling Stimuli

• 1) Thurstone Scaling Converting items into equal
  appearing scale values
• Assumption equal appearing intervals can be
  created from statements or objects based on a
  panel of experts and their opinions.
• We identify the number of scale items and then
  have the experts assign statements (objects) to a
  position on the scale. Discrepancies are
  evaluated and resolved.

http//www.socialresearchmethods.net/kb/scalthur.h
tm
11
Techniques for Scaling Stimuli

• 1) Z Score Converting Ranks to Scale Values for a
  sample of 50 judges

12
Techniques for Scaling Stimuli

• 2) The Semantic Differential measures both the
  intensity and the direction of the attitude.
  (assumes the raw data as interval-scaled)

Respondents are reluctant to give extreme
responses ? may lead to a central tendency error.
Recommendation the two-stage format. Display
Semantic Differential Results as a plot of means
13
Techniques for Scaling Stimuli
3) Stapel Scale an even-numbered, nonverbal
scale used in conjunction with single adjectives
or phrases.
High Quality ( ) 5 ( ) 4 ( ) 3 ( ) 2 ( ) 1 (
) -1 ( ) -2 ( ) -3 ( ) -4 ( ) -5 A Stapel Scale
14
Ratings Scale
15
Techniques for Scaling Respondents

• The Summated scale.
• Format a set of statements to be evaluated on a
  Likert scale.
• Concerns two persons with different attitudes
  may agree on fact central tendency errors.
• The Q-sort technique
• Format each respondent sorts the statements into
  a predetermined number of categories (such as
  most agree with, neutral and least agree
  with ).
• The Differential scale
• Format
• Each statement is assigned a score by
  independent judges
• The respondent selects only the statements he/she
  agrees with
• The average score he/she receives reflects the
  respondents position on the dimension being
  measured.

16
Scaling both Stimuli and Respondents

• Also called response approach to scaling
• Cumulative scales a set of items with which the
  respondent indicates agreement or disagreement.
• A respondents score is calculated by counting
  the number of items answered favorably
• Individuals can be ranked along a unidimensional
  continuum.
• Approach Scalogram analysis

• INSTRUCTIONS Place a check next to each
  statement you agree with.
• ? I believe that this country should allow more
  immigrants in.
• ? I would be comfortable if a new immigrant
  moved next door to me.
• ? I would be comfortable with new immigrants
  moving into my community.
• ? It would be fine with me if new immigrants
  moved onto my block.
• ? I would be comfortable if my child dated a new
  immigrant.
• ? I would permit a child of mine to marry an
  immigrant.

http//www.socialresearchmethods.net/kb/scalgutt.h
tm
17
Multi-item Scales

• Multi-item scale a number of closely related
  individual rating scales whose responses are
  combined into a single index, composite score.
• Used when measuring complex psychological
  constructs.

18
Computer Literacy Scale
19
Multi-item Scales
A Simple Example of a Multi-item Scale to
measure Where Do You Stand Politically Scoring
For each statement, circle A if you Agree, M for
Maybe (or dont know), or D if you
Disagree. COUNT 20 points for every A, 10 points
for every M, and 0 points for every D

• How do you stand on
• PERSONAL ISSUES?
• 20 10 0
• A M D Government should not censor speech,
  press, media or Internet.
• A M D Military service should be voluntary.
  There should be no draft.
• A M D There should be no laws regarding sex
  between consenting adults.
• A M D Repeal laws prohibiting adult possession
  and use of drugs.
• A M D There should be no National ID card.
• SCORING 20 for every A, 10 for every M, and 0
  for every D _______

• How do you stand on
• ECONOMIC ISSUES?
• 20 10 0
• A M D End corporate welfare. No government
  handouts to business.
• A M D End government barriers to international
  free trade.
• A M D Let people control their own retirement
  privatize Social Security.
• A M D Replace government welfare with private
  charity.
• A M D Cut taxes and government spending by 50
  or more.
• SCORING 20 for every A, 10 for every M, and 0
  for every D ______

Source Advocates for Self-Government 269 Market
Place Blvd., 106 Cartersville, GA 30121-2235
20
Multi-item Scales

• LIBERTARIANS support a great deal of liberty and
  freedom of choice in both personal and economic
  matters. They believe governments only purpose
  is to protect people from coercion and violence.
  They value individual responsibility, and
  tolerate economic and social diversity.
• LEFT-LIBERALS generally embrace freedom of choice
  in personal matters, but support central
  decision-making in economics. They want the
  government to help the disadvantaged in the name
  of fairness. Leftists tolerate social diversity,
  but work for what they might describe as
  economic equality.
• RIGHT-CONSERVATIVES favor freedom of choice on
  economic issues, but want official standards in
  personal matters. They tend to support the free
  market, but frequently want the government to
  defend the community from what they see as
  threats to morality or to the traditional family
  structure.
• CENTRISTS favor selective government intervention
  and emphasize what they commonly describe as
  practical solutions to current problems. They
  tend to keep an open mind on political issues.
  Many centrists feel that government serves as a
  check on excessive liberty.
• STATISTS want government to have a great deal of
  control over individuals and society. They
  support centralized planning, and often doubt
  whether liberty and freedom of choice are
  practical options. At the very bottom of the
  chart, left-authoritarians are usually called
  socialists, while right-authoritarians are
  generally called fascists.

Source Advocates for Self-Government 269 Market
Place Blvd., 106 Cartersville, GA 30121-2235
21
Limitations of Scaling Procedures

• It is difficult to measure multidimensional
  stimuli.
• Inability to test individual buyer behavior in
  terms of empirical findings from psychological
  and sociological studies.
• Predictions from attitude scales still need to be
  transformed into measures of more direct interest
  to the marketer.