M303 Attitude Measurement and Scaling
Where we are
Last class
Survey methods and mode choice
Today
Attitude measurement and scaling
Coming next
Questionnaire design
Learning Objectives
You should be able to
Understand the definition and the three components of attitude
Learn the process of measurement and its four steps
Learn four types of measurement scales and specific scales used for each type
Understand the implications of scale selected and selected issues in measurement
Problem
Formulation
Research
Design
Data
Collection
Data Analysis &
Recommendation
Secondary Data
Primary Data
Observational and Qualitative Methods
Surveys
Sampling
Prior-class review
Surveys can collect awareness, perceptions, preferences, motivations, and demographic information.
Telephone interviews typically collect more information per respondent than personal interviews.
Self-administered surveys eliminate interviewer bias but can suffer from low response rates.
A lockbox survey is a single-mode survey that requires no interviewer contact.
What Are Attitudes?
Mental states used by individuals to structure their behaviors.
3 Components of attitude:
Cognitive or Knowledge component
Affective or Liking component
Intention or Action component
3 Components of Attitude: Cognitive or Knowledge Component
A person’s information about an object
Awareness of the existence of an object
Beliefs about the attributes of an object
Judgments about the relative importance of attributes
3 Components of Attitude: Affective or Liking Component
A person’s overall feelings toward an object on a continuum from like to dislike, or favorable to unfavorable
When there are several alternatives, liking is expressed in terms of preference for one alternative
Liking might also be measured by asking which alternative is “most preferred” or “first choice,” which is the “second choice,” and so on
3 Components of Attitude: Intention or Action Component
Refers to a person’s expectations of future behavior toward an object
Intentions are usually limited to a distinct time period
Intentions depend on buying habits and planning horizons
Incorporates information about a respondent’s ability or willingness to pay for the object
Example: Attitude
Organic foods
They use natural fertilizers to produce organic foods.
Organic foods are healthier, but more expensive.
For eggs and dairy items, I prefer organic choices over inorganic ones.
I buy organic baby food.
I buy organic eggs and dairy items.
I do not buy organic fruits and vegetables.
I recommend organic baby food to all parents.
Example: How to Measure the Importance of Price?
Number of times prices are checked
Change in sales when price is changed
Do you check prices regularly?
Do you know the prices in category X?
Do you think price is important?
How important is price?
On a scale of 1 to 10, how important is price?
If the price were 10% higher would you buy this product?
Rank the importance of price, quality and brand name
I consider myself price-sensitive (yes/no)
How to measure?
Process of measurement:
Identify the variable of interest (construct)
Develop a conceptual definition
Develop an operational definition
Choose a measurement scale
Process of Measurement: 1. Identify the variable of interest (construct)
What is a construct?
Specific types of concepts that exist at higher levels of abstraction and are not directly observable but can be measured indirectly through verbal expressions and/or observed behaviors
Examples?
Attitudes
Preferences
Price sensitivity
Brand Loyalty
Intelligence
Process of Measurement: 2: Develop a conceptual definition
What is a conceptual definition?
A description of a construct at the theoretical level, stated in terms of unobservables
Example: Defining Brand Loyalty
Process of Measurement: 3. Develop an operational definition
What is an operational definition?
The translation of a conceptual definition into observable phenomena
A consumer’s emotional attachment to a brand that results in repeat purchase behavior in the present and commitment to purchase in the future.
The degree to which a consumer
Is satisfied with and emotionally attached to her current brand
Repeatedly purchases the same brand
Is resistant to switching despite costs of not doing so
How to Measure… Repeat Purchase?
Behavioral or Attitudinal Measure?
How to Measure… Emotional Attachment?
Behavioral or Attitudinal Measure?
How to Measure… Resistance to Switching?
Behavioral or Attitudinal Measure?
What are the types of measurement scales?
Labels and categories
Labels plus order
Order plus equal intervals
Equal intervals plus a meaningful zero
Types of Scales: Nominal Scales (Categorical)
Classification of variables – mostly expressed in choice or yes/no questions.
Numerical assignment
Coding the categories as 600, 755, and 523 is as valid as coding them as 1, 2, and 3 or A, B, and C
Mathematical Operation
Counting the number of responses falling in each category
Statistical
Can only use mode to measure central tendency
Can be used (after dummy coding) in regression analyses
Types of Scales: Nominal Scale Examples
Yes/No Questions
Choosing a mutually exclusive category
Paired Comparison
Types of Scales: Ordinal Scales
Ranking variables
Numerical Assignment
Coding the ranks as 5, 15, 25, and 40 is as valid as coding them as 1, 2, 3, and 4
Mathematical Operation
counting the number of responses falling in each category
Statistical
Can use both mode and median for central tendency
Can use range for measure of dispersion
People sometimes mistakenly compute the mean.
People mistakenly treat them as continuous in their analyses.
Types of Scales: Ordinal Scale Examples: Rank Order Scales
Respondents rank the objects, factors, or characteristics to indicate their preference
Rank your preferences for the following attributes in making a car purchase decision (1 for the most important and 4 for the least important)
Price -----------
Safety -----------
Design -----------
Fuel economy ------------
Types of Scales: Ordinal Scale Examples: Paired Comparison Scales
Circle the shampoo you prefer in each pair:
Types of Scales: Interval Scales
Rating
Numerically equal distances on the scale represent equal values in the characteristic being measured.
The distance between 2 and 3 is half the distance between 2 and 4
Numerical Assignment
Linear transformations are possible: 1 – 7 can be transformed to 11 – 17
Mathematical Operation
All four mathematical operations can be used
Can’t compare absolute magnitude (“4” ≠ 2 X “2”)
Statistical
Mode, mean, and median as measures of central tendency
Range and standard deviation as measures of dispersion
Hypothesis testing and regression
Types of Scales: Interval Scale Examples: Category Scales
If provided with a coupon, how likely would you be to purchase shampoo X?
Definitely will buy
Probably will buy
Maybe will buy
Probably will NOT buy
Definitely will NOT buy
How often do you try a new shampoo brand?
Always
Most of the time
Sometimes
Never
How confident are you with your shampoo’s quality and performance?
Not at all confident
Neither confident nor unconfident
Very confident
Types of Scales: Interval Scale Examples: Category Scales
________ Very Satisfied
_________ Quite Satisfied
_________ Somewhat Satisfied
_________ Not at all Satisfied
Example 1:
Types of Scales: Interval Scale Examples: Category Scales
Example 2:
Types of Scales: Interval Scale Examples: Category Scales
Example 3:
Types of Scales: Interval Scale Examples: Likert Scales
Respondents indicate their level of agreement with a series of statements on a scale
A Likert item is a statement with which respondents agree or disagree
Uses the same anchors across items
Typically has five response options
Please indicate your agreement with the following statement:
“I always purchase the same brand of shampoo”
Strongly Disagree
Disagree
Neutral
Agree
Strongly Agree
Types of Scales: Interval Scale Examples: Numerical Scales
Respondents select one of the presented scale points
Please rate how important the following attributes are when you purchase shampoo
Would you consider buying a shampoo that is odorless?
Very Unimportant
Very Important
Cost
Ease of Use
Smell
Quality
Brand Name
Definitely NO
Definitely YES
Types of Scales: Interval Scale Examples: Semantic Differential Scales
Five- or seven-point rating scale
Anchored at each end by bipolar adjectives ("sweet-sour");
sometimes monopolar ("sweet-not sweet”)
The respondent marks the blank that best describes the object(s)
A major advantage is that it allows easy comparison on multiple dimensions
Types of Scales: Interval Scale Examples: Semantic Differential Scales
Please evaluate your shampoo brand on the following dimensions
High quality ___ : ___ : ___ : ___ : ___ : ___ : ___ Low quality
Reliable ___ : ___ : ___ : ___ : ___ : ___ : ___ Unreliable
Attractive ___ : ___ : ___ : ___ : ___ : ___ : ___ Unattractive
Fresh ___ : ___ : ___ : ___ : ___ : ___ : ___ Stale
Beneficial ___ : ___ : ___ : ___ : ___ : ___ : ___ Harmful
Warm ___ : ___ : ___ : ___ : ___ : ___ : ___ Cold
Types of Scales: Interval Scale Examples: Comparative Scales
A judgment comparing one object, concept, or person against another
Types of Scales: Interval Scale Examples: Graphical Scales
Types of Scales: Ratio Scales
Rating
Scales with true zero values, where 0 means no allocation or value for the variable
Numerical Assignment
All sorts of transformations are allowable
Mathematical Operation
All four mathematical operations
Statistical
Mode, mean, and median as measures of central tendency
Range and standard deviation as measures of dispersion
Mostly used as a dependent variable in t-tests, regression, ANOVA etc.
Can be used as an independent variable in regression.
Types of Scales: Ratio Scale Examples
What is your annual income before taxes? ___
A natural, unambiguous starting point
We can say that a respondent with an annual income of $40,000 earns twice as much as one with an annual income of $20,000
What is your age?
Types of Scales
Nominal level
Labeling (partition data into categories)
Ordinal level
Labeling + order
Interval level
Labeling + order + equal intervals between points
Ratio level
Labeling + order + equal intervals between points + meaningful zero point
LEVEL OF MEASUREMENT INCREASES FROM
NOMINAL SCALE TO RATIO SCALE
Why does it matter what scale you use?
Answers different questions
Which one of the two cars do you prefer?
Prefer A
Prefer B
Please indicate your preference between the two cars
Prefer A 1 2 3 4 5 6 7 Prefer B
Measurement scales directly determine the types of statistical analyses you can conduct and research questions you can answer.
What are the selected issues in measurement?
Number of scale categories
5? 6? 7? 9? 11? 17?
Strength / Wording of the anchors
like/dislike vs. love/hate
Labeling / Wording of categories
Label them all? Which words?
Monadic vs. Comparative
Monadic
Herbal Essences smells good.
Comparative
Compared to Pantene, Herbal Essences smells better.
What are the selected issues in measurement?
Balance of scales
Balanced
Unbalanced
Forced versus non-forced response choices
Forced
Non-forced
Strongly Disagree
Disagree
Neutral
Agree
Strongly Agree
Strongly Disagree
Disagree
Neither Agree
Nor Disagree
Agree
Strongly Disagree
Disagree
Neutral
Agree
Strongly Agree
N/A
Strongly Disagree
Disagree
Neutral
Agree
Strongly Agree
In-Class Exercise: Brand Perceptions for Microsoft Surface
Conceptual definition
Operational definition
Questions?
Measure the brand—not this particular model.
How to Measure Brand Perceptions for Microsoft Surface?
Conceptual definition
How do the consumers perceive Surface in terms of its personality and attributes
Operational definition
The extent to which consumers think the brand is:
Young
Energetic
Friendly
Funny
Entertaining
Cool
The extent to which consumers evaluate Surface’s attributes favorably:
Flip cover
Keyboard
Durability
Tiles (software)
Microsoft Surface: measurement exercise solution
To sum up
Attitude
3 Components of attitude: Cognitive or Knowledge, Affective or Liking, Intention or Action
Process of measurement
Types of measurement scales
Nominal: yes/no, mutually exclusive categories etc.
Ordinal: Rank order, paired comparison etc.
Interval: Category, numerical, Likert etc.
Ratio
Scale selection and selected issues in measurement
Next class
Survey Design