2. •Agenda
What is Measurement and Scaling?
Types of Measurement Scales
Criteria for Good Measurement
1
2
3
4
Scaling Techniques
3. Measurement
• Researcher must follow rules.
• Reasons for which numbers are usually
assigned:
• Permit statistical analysis for resulting
data.
• Measurement results
What is Measurement and Scaling?
•Introduction
4. Scaling
• Eg: scale satisfaction level from 1 to 11
• Measurement is actual assignment of
numbers
• Scaling is process of placing the respondent
on a continuum.
What is Measurement and Scaling?
•Introduction
6. • Lowest level of measurement.
• Numbers are assigned to the identification of an
objects.
• One-to-one correspondence between the number and
objects.
Nominal Scale
7. • Q. What is your religion?
• 1.Hindusim
• 2.Sikhism
• 3.Christian
• 4.Islam
Number can not be added ,divided ,subtracted or
multiplied.
In Nominal Scale one can compute :-suare Frequency
distribution, Mode, Chi-square test
Example:
8. • Next higher level of measurement.
• The Ordinal scale measurement tells whether an
object has more or less characteristics than other but
it can not tell how much more or less.
• It includes quality ranking, ranking of the teams in
tournament, socio-economic class etc.
• One can compute: Median, percentile, quartiles of the
distribution.
•Ordinal Scale
9. • It takes care of the limitations of ordinal scale.
• There is no absolute zero.
• Difference in 2 adjacent value is constant.
• Technically, numbers on an interval scale can be added
and subtracted but not divided and multiplied.
• All the techniques applicable to the previous scales is
also applicable to interval scale.
•Interval Scale
10. • This is the highest level of measurement takes care of all the
limitations of all other scales.
• Numbers on these scales have meaningful interpretation.
• Ratio scale can be converted into other 3 scales but the vice-
versa is not possible.
• This is the only scale which has a Zero.
•Ratio Scale
6’5” 5’4”5’3” 6’4” 5’11”5’10”
12. • Direct stimulus of object
• Relative in nature
• Non-metric scaling
•Comparative scaling technique
13. Eg: Preference for Toothpaste Brands Using Rank Order
Scaling
• Brand Rank Order
• 1. Crest _________
• 2. Colgate _________
• 3. Close up _________
• 4. Pepsodent _________
• 5. Sensodyne _________
•Rank order comparison
14. Eg: Importance of Bathing Soap Attributes
Using a Constant Sum Scaling
Attribute Points
• 1. Mildness 8
• 2. Lather 2
• 3. Shrinkage 3
• 4. Price 53
• 5. Fragrance 9
• 6. Packaging 7
• 7. Moisturizing 5
• 8. Cleaning Power 13
Sum 100
•Constant Sum Scaling
15. Eg: Obtaining Shampoo Preferences Using Paired
Comparisons
Recording form: Head & shoulder Clinic plus Dove
Head & shoulder 0 1
Clinic plus 1 1
Dove 0 0
No. of times 1 0 2
preferred
•Paired Comparison
16. • It discriminates among large number of objects
• Ordinal in nature
• Time consuming
• Expensive
•Q-sort Technique
18. •Continuous rating
Allows the respondent to place a mark at any point among a line
running between two extreme points
Also called graphic rating scale.
Has metric scaling
It has interval or ratio scale
Can be expressed using diagrams, smiley and scale
19.
20.
21. •Itemized Rating
In this respondents are
provided with a scale that
has a number of brief
descriptions associated
with each response
categories.
Issues to be kept in mind :
•No of catergories to be used
•Odd or even categories
•Balanced v/s unbalanced scales
•Nature and degree of verbal
description
•Forced v/s non forced scales
•Physical form
It includes
•Likert
•Semantic diffrentiation
•Stapel
22.
23. •Likert scale
Named after Mr. Rensis Likert
Range of categories from strongly disagree to strongly
agree
Also called summated scale
It takes longer to complete
24. •Semantic Differentiation
Used to compare images of competing
brands or services.
Respondent is required to rate each
attitude on a number of 5- 7 pt rating
scale
+ve and –ve descriptions can simetimes
switch places - reduces tendency to just
mark 1 side
25.
26. •Stapel Scaling
Ian stapel developed it
Its presented vertically
with an adjective in
the middle and five
data points abv and
five data point below.
No zero point is
offered.
29. • It refers to the whether we are measuring what
we want to.
•Validity
Content
• Subjective Judgment for assessing the appropriateness of
the construct.
Concurrent
• Measures the validity of new measuring techniques by
correlating them to the established techniques.
Predictive
• The ability of one phenomenon at that point of time to
measure another phenomenon at a future point of time.
30. •Reliability
Its concerned with consistency, accuracy and
predictability of the scale.
It refers to the extent to which a measurement
process is free from random errors.
I. Test –retest II. Split-Half
31. •Sensitivity
It refers to
instrument’s ability
to measure the
variability in a
concept.
Sensitivity can be
increased by adding
questions or items
to the scale.
Instead of only
Agree and Disagree
we can add strongly
agree and strongly
disagree to the scale