Definition of Measurement

1
Definition of Measurement

• Measurement is a procedure used to assign
  numbers that reflect the amount of an attribute
  possessed by an event, person, or object.
• We measure attributes, not the person or event

2
The Measurement Process

• What is a construct?
• Specific types of concepts which exist at high
  level of abstraction, usually theoretical
• They are generally not observable but are
  inferred through indirect means
• example brand loyalty

3
The Measurement Process

• A theoretical definition of a construct
• States the central idea or concept
• Defines the concept in terms of other concepts
  and constructs
• Establishes boundaries for the construct

4
The Measurement Process

• An operational definition
• Defines which observable characteristics will be
  measured
• Defines the process for assigning a value to the
  concept
• So the construct is broken down into measurable
  components and the method for creating an overall
  value based on those components is specified

5
Construct Frustration with the Purchase Process

• Theoretical Definition Blocking Goals
• an imposed delay of reward operationalized by an
  interruption or blocking of a consumer's progress
  towards a valued goal
• Operational Definition
• Used focus groups to develop 45 items
• Analysis resulted 12 items which defined two
  dimensions
• Frustration with processing information
• Frustration with the pace of technology
• Measurement Scale (Agree/Disagree 7 pt. Scale)
• Searching for relevant information about which
  type of computer to purchase is a very
  frustrating process.
• I can not decide when to buy a computer because
  it seems as if there is always a better
  technology just a few months away.

6
Example of the measurement process

• Theoretical Definition Role Ambiguity
• discrepancy between the information available to
  the person and that which is required for
  adequate performance
• Operational Definition Amount of Uncertainty
• uncertainty an individual feels regarding job
  role responsibilities from other employees and
  customers
• Measurement Scale
• scale consists of 45-items -- each item assessed
  by a five-point scale with category labels 1
  very certain, 2 certain, 3 neutral, 4
  uncertain, 5 very uncertain.
• How much freedom of action I am expected to have
• How I am expected to handle non-routine
  activities on the job
• The sheer amount of work I am expected to do

7
The Measurement Process
Research findings
Develop a construct
Utilizes the scale
Identify the concept of interest
Create a theoretical definition
Evaluate the reliability validity of scale
Create an operational definition
Measurement Scale
8
The Measurement Process

• Developing measurement scales
• Once a construct has been defined (both
  theoretically and operationally) the researcher
  begins to create a scale that will be used to
  measure the construct in subsequent research
• There are four different types of scales
• Nominal
• Ordinal
• Interval
• Ratio

9
Example Mode, Median, Mean
FREQUENCY DISTRIBUTION OF FAMILIARITY WITH THE
INTERNET
10
Nominal Scales

• Characteristics
• Gives a name or label to a category
• Doesnt show rank order
• Doesnt have equal intervals
• Examples
• What is your gender?
• 1. Male
• 2. Female
• Where did you last buy toothpaste?
• 1. In a supermarket
• 2. In a discount store
• 3. Somewhere else

11
Operations with Nominal Scales
400 Males 600 Females 1000 Total

• What can we say?
• 40 are male and 60 are female
• there are more females than males
• We can only do two things
• calculate percentages
• count which category is the largest.
• This is the modal response

12
Ordinal Scales

• Characteristics
• Numbers indicate the relative positions of the
  objects but not the magnitude of differences
  between them
• Measures go from highest to lowest most to
  least strongest to weakest etc.
• Doesnt have equal intervals
• Example How many movies did you see last month?
• 1. None
• 2. One or two
• 3. Three or more

13
Operations with Ordinal Scales
Number of Movies Frequency 1. None 40
80 2. One or two 30 60 3. Three or
more 30 60 100 200
What can calculate with an ordinal scale? Mode
- most frequently occurring response NONE
(category 1) Median - 50th percentile
response One or two (category 2)
14
Interval Scales

• Characteristics
• Naming
• Ranking
• Equal Intervals
• Example How likely are you to buy a new
• Ford Voyager this year?
• 1. Very likely
• 2. Likely
• 3. Neither likely or unlikely
• 4. Unlikely
• 5. Very unlikely

15
Operations with Interval Scales
Likelihood Frequency Percent 1 10 10
2 10 10 3 40 40 4 25 25 5 15
15
What can calculate with an interval scale? Mode
- most frequently occurring response (3,
40) Median - 50th percentile response (3) M
ean - average of the distribution (4.25)
16
Ratio Scales

• Characteristics
• Naming
• Ranking
• Equal Intervals
• True zero point
• Example
• How old are you?
• How many people are in your class?
• How many miles do you travel to school?

17
Operations with Ratio Scales
What can calculate with an ratio scale? All
mathematical operations What can we measure with
ratio scales? All variables that are
countable and have a true zero
point Examples amount of the GNP your weight
18
One Variable, All Measurement Levels
Many variables can be measured at any level
of measurement. Example Age Nominal Ordinal
Are you older then 30? How old are you?
Yes 40
Interval Ratio How old are you? How old are
you? 21-30 31-40
41-50 51-60
19
Number of times people ate dinner at a restaurant
last week... You sample 100 people
Nominal Did you eat out at a restaurant last
week? Yes No 80 20 Ordinal Circle
correct answer regarding the of times . 0
20 Once 20 Twice or
more 60
20
Interval Circle correct answer
Frequency 1. 0-1 60 2. 2-3 20 3. 4-5 1
0 4. 6-7 10
(60 .5) (202.5) (104.5) (106.5) 100
1.9
21
Ratio How many times did you eat at a
restaurant last week? (fill in )
Frequency
Response
0 20 020 0 1 30 130 30 2 10 210
20 3 20 60 4 6 24 5 4 20 6 7 42 7
3 21 217/100 2.17
22
The Measurement Process

• Sources of measurement differences
• A true difference in the characteristic being
  measured
• example in a study comparing two teaching
  techniques, group A learned more than group B
• Differences due to stable characteristics of
  individual responses
• example group A happens to be more intelligent
  than B
• Differences due to short-term personal factors
• example group A was in a better mood than group B

23
The Measurement Process

• Sources of measurement differences cont.
• Differences caused by situational factors in the
  interview situation
• example group B was tested on a hot day without
  air conditioning
• Differences resulting from variation in
  administering the survey/experiment
• example the experimenter for group A was nicer
  than the experimenter for group B

24
The Measurement Process

• Sources of measurement differences cont.
• Differences due to the sampling of items included
  in the questionnaire
• example the measure of learning only included
  memorization, not analytical skills
• Differences due to a lack of clarity in the
  measurement instrument
• example the test questions were confusing
• Differences due to mechanical or instrument
  factors
• example the test had missing questions

25
The Measurement Process

• Back to reliability and validity
• All measurement contains error
• M A E
• where M is the measurement
• A is the true construct and E is error
• This error can be random or systematic
• Researchers try to reduce the amount of error so
  that they can get more accurate differences
  reflected in the measure

26
Reliability

• Reliability
• A measurement scale that provides consistent
  results over time.
• Example Ruler
• Internal Consistency Reliability
• A measure that assesses the ability to produce
  the similar results using different samples to
  measure a phenomenon during the same time period.
• Cronbach-Alpha Correlation

27
Validity

• Validity
• Are we measuring what we set out to measure?
• Example
• We want to know if people will buy our new brand
  of champagne.
• We ask people how they like it.
• We havent measured purchase intention we have
  measured attitude toward the product!
• OUR MEASURE IS NOT VALID

28
Validity Different Perspectives

• Face
• The degree to which a measurement seems to
  measure what it is supposed to measure
• Content
• The degree to which measurement items represent
  the universe of the concept under study
• Criterion-related
• The degree to which a measurement instrument can
  predict a variable that is designated a criterion
• Construct
• The degree to which a measurement instrument
  confirms a hypothesis created from a theory based
  on the concepts under study
• Convergent
• Discriminant

29
Reliability vs. Validity
How Reliability and Validity Interact

Neither reliable nor valid
Reliable, but not valid
Both reliable and valid
30
Definition of Attitudes

• An attitude is an enduring organization of
  motivational
• emotional
• perceptual
• or cognitive processes
• with respect to some aspect of our environment.

It is a learned predisposition to respond in a
consistently favorable or unfavorable manner
toward an object.
31
The Relationship between Attitudes and Behavior

• The link between attitudes and behavior is very
  complex
• Typically, we think of a progression

• Better when make assertions about groups rather
  than individuals
• Attitudes based on actual experience are more
  highly correlated with behavior than attitudes
  based on advertising exposure

32
Changing Attitudes

• Based on the (Belief/Importance Model) of
  attitudes

where the overall evaluation of a brand is a
function of B beliefs about the attributes
possessed by the brand I the importance of
each attribute (Note the Fishbein Model
belief/evaluation model based on behavior
learning theory)

• There are three ways to change attitudes
• Change peoples beliefs about attributes
• Change the relative importance of the attribute
• Add new beliefs

33
Changing Attitudes

• Attitude S ImportanceAttribute Belief
• Example Attitude about Hondas
• Attribute Style
• Importance High Belief Positive
• Attribute MPG
• Importance Medium Belief Positive
• Attribute Made in America
• Importance Low Belief Negative
• Total Favorable attitude towards Hondas

34
The Relationship between Attitudes and Behavior

• Factors that reduce the link between attitudes
  and behavior
• Needs or motivations
• Ability (e.g., to purchase)
• Trade-offs within between product categories
• Increased information prior to behavior
• Incomplete measurement

35
Attitude Scales Defined

• Scaling refers to techniques to determine
  quantitative measures of subjective and abstract
  concepts.
• A Scale is a measurement tool.
• Once we have numbers to represent concepts, we
  can make comparisons, e.g., correlations

36
Attitude Scales

• Creating scales to measure attitudes is
  difficult due to the fact that attitudes are not
  directly observable

37
Attitude Scales

• The first scaling issue
• Are we measuring a single global construct
  (unidimensional scale)?
• or several dimensions of a construct
• (multidimensional scale)?
• Note ?
• Many questions can be asked in a unidimen-sional
  scale, but they are combined to form one
  construct (e.g., whether swimmers believe Turbo
  suits are fast)

38
Turbo Survey

• 1. Turbo swimsuits help a swimmer swim fast.
• Disagree 1 2 3 4 5 Agree
• 2. Turbo swimsuits are used by elite swimmers.
• Disagree 1 2 3 4 5 Agree
• 3. Turbo swimsuits are not for fast swimmers.
  (reverse scored)
• Disagree 1 2 3 4 5 Agree

39
Do swimmers believe Turbo suits are fast?
If these three items (questions 1-3) are
correlated, then we would take the average of the
three questions. This average is 4.67. Item 3
is reverse scored This number represents the
subjects attitude towards how fast Turbo suits
are.
40
Graphic Rating Scales
Overall, how would you rate the taste of Pepsi
One? Excellent Fair
Very Poor
Numerical data obtained from the scales is
typically treated as interval data
41
Graphic Rating Scales
Examples
Overall, how would you rate the taste of Pepsi
One? Excellent Very Poor
0 10 20 30
40 50 60 70
80 90 100
42
Itemized Rating Scale
Similar to a graphic rating scale, but with fewer
categories. Simplifies respondents task, so
results are often more reliable.
How would you rate Levis jeans? Excellent
Good Fair Poor Durability (
) ( ) ( ) ( )
Value ( ) ( )
( ) ( ) Fashion ( )
( ) ( ) ( )
43
Rank-Order Scale
Example
lifelike good
good appearance huggability
size Mattel ( ) (
) ( ) Fisher-Price ( )
( ) ( )
Calico ( ) ( )
( )
Please rank the following frog toys with 1 being
the brand that best meets the characteristics
being evaluated and 3 being the worst brand on
that characteristic.
44
Q-Sorting
Form of rank ordering. Set of objects is given
to respondent to sort into rating
categories. Example cards with different
potential features that could be designed into a
software product. Excellent
Poor Feature
Feature 10 9 8 7 6 5
4 3 2 1 0 3 4 7
10 13 16 13 10 7 4 3
45
Constant Sum Scales
Respondent divides a given number of points
among various attributes of a product, based on
the perceived importance of the
attributes. Example Below are five attributes of
hair spray. Please allocate 100 points among the
attributes so that you give the attribute that is
most important to you, the most
points. Attributes Points Not sticky (
) Even-releasing spray ( ) Good
Value ( ) Pleasing fragrance (
) Strong holding power ( ) TOTAL
POINTS 100
46
Semantic Differential Scale
Select bipolar words or phrases to describe a
concept on a scale with 4 to 7 points. Means for
each pair of adjectives are computed and an image
profile is created.
Strawberries are Mean of each
adjective pair 1 2 3 4 5 6
7 Low calorie ( ) ( ) ( ) ( ) ( ) (
) ( ) High calorie Special fruit ( ) (
) ( ) ( ) ( ) ( ) ( ) Everyday
fruit Nutritious ( ) ( ) ( ) ( ) ( ) (
) ( ) Empty calories Bad-tasting ( ) (
) ( ) ( ) ( ) ( ) ( ) Delicious
47
Likert Scales

• Likert Scales consists of a series of statements
  that express either a favorable or unfavorable
  attitude toward a concept
• Example Going to amusement parks is one of my
  favorite activities.
• ( ) Strongly agree
• ( ) Agree
• ( ) Neither agree or disagree
• ( ) Disagree
• ( ) Strongly disagree

48
Purchase Intent Scales
Example
If this coffee maker sold for approximately 75
and were available in the stores where you
normally shop, how likely would you be to buy
it? ( ) definitely will ( ) probably will (
) might or might not ( ) probably will not (
) definitely will not Most commonly-used
marketing research scale. When combined with
estimates about actual purchase likelihood can be
the basis for go-no-go decisions.
49
Type of Scale Selection

• Considerations for which type of scale to use
• Rating vs. ranking vs. purchase intent
• Ease of use, needs of client
• Balanced vs. non-balanced
• Number of categories
• Most range between five and nine categories
• Odd or even number of scale categories
• Neutral point?
• Forced vs. non-forced choice
• Dont know?