You are on page 1of 18

Scaling Methods

24 June 2011

Prepared by Prof C Y Nimkar

Scaling Methods
 Two types of scales


Comparative Scales


Compare one object with another Evaluate each object independently

Non-comparative scales


24 June 2011

Prepared by Prof C Y Nimkar

Comparative Scales

24 June 2011

Prepared by Prof C Y Nimkar

Comparative scales
 Shows relative importance between two

objects  Relatively small differences are detected

24 June 2011

Prepared by Prof C Y Nimkar

1. Paired comparison
 Respondent is shown a pair of objects  He is asked to select one object that he

prefers

24 June 2011

Prepared by Prof C Y Nimkar

1. Paired comparison Example 1


Que.: I will show pair of brands. Please tell me which brand do you prefer from each pair Brand 1 LG LG Akai Samsung Brand 2 Samsung Akai Sony Sony Preferred

24 June 2011

Prepared by Prof C Y Nimkar

1. Paired comparison Example 1


Preferred
LG Samsung Akai Sony Total

LG
0.00 0.80 0.25 0.60 1.65

Samsung
0.20 0.00 0.10 0.30 0.60

Akai
0.75 0.90 0.00 0.65 2.30

Sony
0.40 0.70 0.35 0.00 1.45

24 June 2011

Prepared by Prof C Y Nimkar

1. Paired comparison Example 2


Que.: Allocate a total of 11 to brands shown in each pair
Brand 1 LG LG Akai Samsung Brand 2 Samsung Akai Sony Sony Score to brand Score to brand 1 2

24 June 2011

Prepared by Prof C Y Nimkar

1. Paired comparison
 Easy to administer.  Not suitable if number of pairs is large.

24 June 2011

Prepared by Prof C Y Nimkar

2. Constant sum scale


 Respondents are shown all objects to be

compared at one time.  He is asked to allocate points to each object out of total points say 100.

24 June 2011

Prepared by Prof C Y Nimkar

10

2. Constant sum scale - Example


Que.: Below are five attributes that you might consider while buying a 4wheeler. Please allocate a total of 100 points to them. More the points, more important is the attribute to you. You can allocate 0 point if the attribute is not at all important to you. Attribute Mileage Seating capacity Luggage space Proximity of dealer Servicing network TOTAL 100 Points

24 June 2011

Prepared by Prof C Y Nimkar

11

2. Constant sum scale


 Data gives relative importance.  Information on intensity is obtained.  Faster than paired comparison.  Data is either in interval scale or ratio scale.

24 June 2011

Prepared by Prof C Y Nimkar

12

Noncomparative Scales

24 June 2011

Prepared by Prof C Y Nimkar

13

1. Continuous rating scale


 Respondent is asked to mark (X) anywhere

on a line.
Que.: Please indicate your opinion about work of NGOs by placing (X) at the appropriate location on the line shown below

Unfavourable

Favourable

24 June 2011

Prepared by Prof C Y Nimkar

14

2. Itemised rating scale


 Respondent is asked to select one category.
Que.: Please indicate your opinion about work of NGOs by placing (X) in one of the categories shown below.
Unfavourable Favourable

Extremely

Quite

Slightly

Neither

Slightly

Quite

Extremely

24 June 2011

Prepared by Prof C Y Nimkar

15

2. Itemised rating scale


 No optimal number of categories.  Depends on respondents ability to

discriminate among categories.  Generally 5 to 9 categories are included.  5 points rating scale is popular.  Unbalanced categories when responses are skewed.  Separate category for no opinion is desirable.
24 June 2011 Prepared by Prof C Y Nimkar 16

2. Itemised rating scale


5 points Definitely Probably Might or might not Probably not Definitely not Can not say 7 points Definitely Mostly Probably Might or might not Probably not Mostly not Definitely not Can not say 9 points Definitely Very likely Likely Fairly likely Might or might not Fairly unlikely Not likely Almost not Definitely not Can not say

24 June 2011

Prepared by Prof C Y Nimkar

17

2. Graphic itemised rating scale


 Respondents are shown pictures

 Used for children.

24 June 2011

Prepared by Prof C Y Nimkar

18

You might also like