Transcript Measurement in Survey Research

Measurement in Survey
Research
MKTG 3342
Fall 2008
Professor Edward Fox
Measurement in Survey Research
Measurement is the process of assigning
numbers or labels to the attributes of
objects, persons, states, or events in
accordance with specific rules
The Measurement Process
Research
findings
…which leads
to…
Utilize the
scale
If the evaluation is
satisfactory, the
researcher…
Evaluate the
reliability and
validity of the
scale
…that requires the
researcher to…
Develop
a construct
Identify the
concept of
interest
…which is
used to…
A measurement
scale
…which enables a
researcher to create…
…which is used
to create…
A constitutive
definition
An operational
definition
…which enables
a researcher to
develop …
Steps 1&2: Identify Concept /
Develop Construct
Research
findings
Utilize the
scale
Evaluate the
reliability and
validity of the
scale
Develop
a construct
Identify the
concept of
interest
A measurement
scale
A constitutive
definition
An operational
definition
Steps 1&2: Identify Concept /
Develop Construct
 Measurement
begins by identifying a
concept of interest and the construct
to be studied. Both are abstractions of
reality.
 A concept is expressed in every-day
terminology. This requires the
researcher to generalize/categorize.
 A construct is a theoretical abstraction
that can’t really be observed (e.g.,
love, trust, social class, personality,
power).
Steps 1&2: Identify Concept /
Develop Construct – Example

Question – Why do some customers
buy Air Jordan athletic shoes over and
over again?

Concept – Repeat purchase

Construct – Brand loyalty
Steps 3&4: Define the Concept
Conceptually and Operationally
Research
findings
Utilize the
scale
Evaluate the
reliability and
validity of the
scale
Develop
a construct
Identify the
concept of
interest
A measurement
scale
A constitutive
definition
An operational
definition
Steps 3&4: Define the Concept
Conceptually and Operationally
Constitutive Definition
A
theoretical or conceptual definition that
defines the concept in terms of other
concepts and constructs; like a dictionary
definition
Operational Definition
 Defines
which observable characteristics will
be measured and the process for assigning a
value to the concept
Steps 3&4: Define the Concept
Conceptually and Operationally – Example

Constitutive Definition – Increased propensity
to purchase a brand due to previous
experience with that brand

Operational Definition – Rating of purchase
probability, depending upon prior purchase
Step 5: Develop a Measurement
Scale
Research
findings
Utilize the
scale
Evaluate the
reliability and
validity of the
scale
Develop
a construct
Identify the
concept of
interest
A measurement
Scale
A constitutive
definition
An operational
definition
Step 5: Develop a Measurement
Scale
SCALE
 A scale is a set of symbols or numbers so
constructed that the symbols or numbers can
be assigned by a rule for the individuals (or
their behaviors or attitudes) to whom the
scale is applied
Types of Scales
 Nominal
 Description
 Uses
numerals to identify objects, individuals,
events, or groups. Used for Classification
(male/female; buyer/nonbuyer)
 Typical
Descriptive Statistics
 Frequency
 Examples


counts, percentages/modes
of Nominal Scales
Gender
Geographic Area
(1) Male
(1) Urban
(3) Suburban
(2) Female
(2) Rural
Types of Scales
 Ordinal
Scale
 Description
 In
addition to identification, the numerals
provide information about the relative amount
of some characteristic; determines greater or
less than
 Typical
Descriptive Statistics
 Median
 Example

of Ordinal Scale:
Please rank the following fax machines from 1 to 5 with 1 being
the most preferred and 5 the least preferred.
 _____ Panasonic
 _____ Toshiba
 _____ Sharp
 _____ Savin
 _____ Ricoh
Types of Scales
 Interval
 Description
 Has
all the properties of nominal and ordinal
scales + equal intervals between consecutive
points; preferred measure for complex
concepts or constructs
 Typical
Descriptive Statistics
 Mean/variance
 Example
 Scaled
of an Interval Scale
response (on a scale from 1 to 10…)
Types of Scales
 Ratio
 Description
 Incorporates
all the properties of nominal,
ordinal, and interval scales plus it includes an
absolute zero point
 Typical
Descriptive Statistics
 Mean/variance
 Example
 Age,
+ a few higher order statistics
of a Ratio Scale
weight, height, population of the U.S., etc.
Step 5: Develop a Measurement
Scale – Example

Probability of purchasing the shoe
brand again in the next month, next
year, or at any point in the future

Type of scale?
Step 6: Scale Reliability and
Validity
Research
findings
Utilize the
scale
Evaluate the
reliability and
validity of the
scale
Develop
a construct
Identify the
concept of
interest
A measurement
scale
A constitutive
definition
An operational
definition
Step 6: Scale Reliability and
Validity
 Any
measurement can be expressed as a
function of three components:
XO = XT + XS + XR
Observed Score = True Score + Systematic Error + Random Error
 Ideally,
In Practice,
XO = XT
XO  XT …
that is, XS + XR  0
Total Error = XS + XR, where
XS = Systematic error (validity)
XR = Random error (reliability)
Step 6: Scale Reliability and
Validity – Illustration
Not Reliable
.
.
.
.
.
.
Not Valid
.
. .
.
.......
..
......
.
.
.
Reliable and Valid
Scale Reliability
The degree to which measures are free
from random “noise” and, therefore,
provide consistent data
 Issues
 Test-Retest Reliability
 Internal Reliability (split-half
technique)

Scale Validity
Addresses the issue of whether what we
attempted to measure was actually
measured
 Issues

 Face
Validity
 Content Validity
 Predictive Validity
 Convergent Validity
 Discriminant Validity
Summary
Measurement means using rules to assign
numbers to objects in such a way as to
represent quantities of attributes
 The measurement process is as follows:
identify the concept of interest, develop a
construct, define the concept constitutively
and operationally, develop a measurement
scale, evaluate the reliability and validity of
the scale, and then use the scale
 There are four basic types of measurement
scales: nominal, ordinal, interval, and ratio

Summary
(Cont.)
Measurement data consists of information
(“signal”) and error (“noise”).
 Validity requires that you actually measure
what you intended to measure
 Reliability is the degree to which measures
are free from random error
