0% found this document useful (0 votes)
113 views28 pages

Measurement and Scaling Techniques

This document discusses measurement scales and scaling techniques. It describes the four primary measurement scales: nominal, ordinal, interval, and ratio scales. It also covers comparative scaling techniques like paired comparison and rank ordering as well as non-comparative techniques like Likert scales. Finally, it discusses evaluating the reliability, validity, and generalizability of multi-item measurement scales.

Uploaded by

Riya Pandey
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
113 views28 pages

Measurement and Scaling Techniques

This document discusses measurement scales and scaling techniques. It describes the four primary measurement scales: nominal, ordinal, interval, and ratio scales. It also covers comparative scaling techniques like paired comparison and rank ordering as well as non-comparative techniques like Likert scales. Finally, it discusses evaluating the reliability, validity, and generalizability of multi-item measurement scales.

Uploaded by

Riya Pandey
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 28

MEASUREMENT and

SCALING TECHNIQUES
MEASUREMENT
• Process of assigning meaningful and
relevant numbers or symbols to the
characteristics of objects according to
certain pre-specified rules.
• Level of measurement is a function of
rules under which numbers are assigned
• One-to-one correspondence between the
numbers and the characteristics being
measured.
MEASUREMENT SCALES
• Four Primary Scales of Measurement:
– NOMINAL SCALE
– ORDINAL SCALE
– INTERVAL SCALE
– RATIO SCALE
NOMINAL SCALE
• Numbers serve only as labels or tags for
identifying and classifying objects.
• Numerical symbols for nomenclature
purpose only.
• Strict one-to-one correspondence.
• Numbers do not reflect the amount of
characteristics possessed by the objects.
• Simplest and least powerful.
NOMINAL SCALE
• Only permissible operation – counting
(frequency counts and percentages).
• Measure of Central Tendency- Mode.
• Measure of Dispersion- Nil.
• Test of significance- Chi-square Test
ORDINAL SCALE
• Ranking or ordered scale of lowest order.
• Numbers are assigned to indicate the
relative extent to which some
characteristics is being possessed.
• Indicative of order (relative position) but
not the magnitude of difference between
two objects.
• Equivalent objects receive same rank.
ORDINAL SCALE
• Permissible operations- Counting and
statistics based of centiles.
• Measure of Central Tendency- Median.
• Measure of Dispersion- Quartile Deviation.
• Measure of Relationship- Rank-Order
Correlation.
• Test of significance- Non-Parametric
Tests.
INTERVAL SCALE
• Numbers are used to rate objects such that
numerically equal distances on the scale
represent equal distances in the characteristics
being measured.
• Contains all information of an ordinal scale but
also allows comparison of distances.
• Zero-point and the unit of measurement is
arbitrary.
• Linear transformation is allowed.
• Not meaningful to take ratios because the zero-
point is not fixed.
INTERVAL SCALE
• Permissible operations- Counting, ordering
and measurement.
• Measure of Central Tendency- A. M.
• Measure of Dispersion- Range and
Standard Deviation.
• Measure of Relationship- Product-Moment
Correlation.
• Test of significance- Parametric Tests,
Regression.
RATIO SCALE
• It has the power of all the preceding scales
as also the concept of absolute-zero.
• Allows identification/classification, ranking,
comparing intervals/differences, also
computation of meaningful ratios.
• Allows only proportionate transformation of
the form y = bx.
• Most powerful, represents the actual
amount of measurement.
• All statistical techniques can be applied.
SCALING
• Extension of measurement.
• Involves creation of a continuum upon
which measured objects are located.
• Scaling techniques help to measure
abstract concepts more accurately.
• Scaling Techniques:
– Comparative Scales
– Noncomparative Scales.
COMPARATIVE SCALES
• Involves the direct comparison of the
stimulus objects.
• May be interpreted in relative terms and
have only Ordinal or Rank-Order
properties.
• Non-metric scaling.
• Advantages:
– Small differences can be detected.
– Easily understood and applicable.
– Reduced halo/carryover effect.
COMPARATIVE SCALES
• Disadvantages:
– Ordinal nature of the data.
– Inability to generalize beyond stimulus objects
scaled.
• Comparative Scaling Techniques:
– Paired Comparison Scaling
– Rank Order Scaling
– Constant Sum Scaling
COMPARATIVE SCALES:
Paired Comparison Scaling
• Respondent is presented with two objects
and asked to select one according to some
criterion.
• Data obtained is ordinal.
• Most widely used comparative scaling
technique.
• For n brands nC2 comparisons are required
• Useful when number of objects/brands are
limited.
COMPARATIVE SCALES:
RANK ORDER SCALING
• The second most popular technique.
• Respondents are presented with several objects
simultaneously and asked to order/rank them
according to some criterion.
• Commonly used to measure preferences for
brands as well as attributes.
• Advantages - More closely resembles shopping
environment, takes less time, only (n-1) scaling
decisions, easily understandable.
• Disadvantage- Only Ordinal Data.
COMPARATIVE SCALES:
CONSTANT SUM SCALING
• Respondents are required to allocate a
constant sum of units among a set of stimulus
objects with respect to some criterion.
• Allocation of units reflects the importance
attached to each attribute.
• Some times treated as metric but strictly
speaking they are ordinal scale because of its
comparative nature and lack of
generalizability.
• Advantage- allows for fine discrimination.
• Disadvantages- process is cumbersome.
NON-COMPARATIVE SCALES
• Monadic (one object at a time) or Metric Scales.
• Each stimulus object is scaled independently of
others in the stimulus set.
• Interval or Ratio Scaled, can be generalized.
• Non-Comparative Scales:
– Continuous Rating Scales.
– Itemized Rating Scales:
• Likert Scales
• Semantic Differential Scales.
• Stapel Scales.
NON-COMPARATIVE SCALES:
CONTINUOUS RATING SCALE
• Graphic rating scale.
• Respondent has to rate the objects by placing a
mark at the appropriate position on a line running
from one extreme of the criterion to the other.
• Forms: horizontal/vertical, scale points, in form of
numbers/brief descriptions.
• Scores are treated as Interval Data.
• Easy to construct, freedom to choose.
• Scoring is cumbersome and unreliable.
NON-COMPARATIVE SCALES:
ITEMIZED RATING SCALES
• Scale having numbers and/or brief
description for each category.
• Categories are ordered in terms of scale
position.
• Interval Data.
• Most widely used.
• Types:
– Likert Scale
– Semantic Differential Scale
– Stapel Scale
NON-COMPARATIVE SCALES:
ITEMIZED RATING SCALES
• LIKERT SACLE:
– Scale with 5 response categories ranging
from “strongly disagree” to “strongly agree”.
– Numerical Score: 1 to 5 or -2 to +2.
– Analysis on item-to-item basis (profile
analysis) or total (summated) score.
– Consistent scoring procedure for summated
approach.
– Easy to construct and administer, easily
understandable.
NON-COMPARATIVE SCALES:
ITEMIZED RATING SCALES
• SEMANTIC DIFFERENTIAL SCALE:
– 7-point rating scale with endpoints associated
with bipolar labels that have semantic
meaning.
– Mixing of positive and negative
adjectives/phrases to control response-inertia.
– Numerical Score: 1 to 7 or -3 to +3.
– Data analysis by “Profile Analysis”, using
mean/median followed by statistical analysis.
– Very versatile: comparing brands, products,
image, develop advertising & promotional
strategies, NPD.
NON-COMPARATIVE SCALES:
ITEMIZED RATING SCALES
• STAPEL SCALE:
– Unipolar rating scale with 10 categories
(even) numbered from -5 to +5 without a
neutral point.
– Single adjective in the middle.
– Attitude measurement.
– Analysis is same.
– Not very popular - confusing and difficult to
apply.
NON-COMPARATIVE ITEMIZED
RATING SCALE DECISIONS
• Number of Scale Categories
• Balanced vs Unbalanced Scales
• Odd or Even Number of Categories
• Forced vs Non-forced Scales
• Nature & Degree of Verbal Description
• Physical Form or Configuration
MULTIITEM SCALE EVALUAION
• A multi-item scale should be evaluated for
accuracy and applicability.
• Assessment of Reliability, Validity and
Generalizability.
MEASUREMENT ACCURACY
• Measurement Error: Variation in the
information sought (by the researcher) and
information generated (by the
measurement process).
• True Score Model:
XO = XT + XS + XR
– Systematic Error (XS) affect the measurement
in a constant way (stable factor) but Random
Error (XR) is not constant.
RELIABILITY
• Extent to which a scale produces
consistent results if repeated
measurements are made.
• Random Error produces inconsistency,
leading to lower reliability.
• Assessed by determining the proportion of
systematic variation in a scale, i.e. by
determining the association between
scores obtained by different administration
of scale.
RELIABILITY
Methods of Assessing Reliability
• Test-Retest Reliability: Identical sets of scale
items at two different times and correlated.
• Alternative-Forms Reliability: Two equivalent
forms of scale at two different times and correlated.
• Internal Consistency Reliability: To assess
internal consistency of set of several items
(summated scale).
– Split-Half Reliability: Items are divided into 2 halves
and the resulting half scores are correlated.
– Coefficient alpha (Cronbach’s alpha): Average of all
possible split-half coefficients resulting from different
ways of splitting.
VALIDITY
• The ability to scale or measuring instrument to
measure what is intended to measure (No
measurement error).
• Content Validity:
– Face validity.
– Subjective & systematic evaluation of the
representativeness of the content of a scale (measure
what is intended to measure).
• Criterion Validity:
– Ability to correlate with other measures of the same
construct.
• Construct Validity:
– Empirical evidences generated by a measure is
consistent with the theoretical logic about the concept.

You might also like