CLUSTERING METHODS 2 TYPES

1ST HIERARCHICAL CLUSTERING : NO OF CLUSTER CAN BE

FORMULATED.
STEPS
DATA VIEW >> ANALYZE>> CLASIFY>> HIERARCHICAL
CLUSTER>>
2ND K- MEANS (NON- HIERARCHICAL)

Q.1. Identify the number of clusters / segments from the

data provided.
Q.2. Determine which respondent belongs to which
segment.
Q.3. Identify the variables that distinguish the segments at
90% confidence level.
Q.4. Create segment profiles for each cluster based on their
scores on the distinguishing variables.
Q.5. Label / Name each segment based on their dominant
distinguishing characteristics.
OUTPUT
HIERARCHICAL CLUSTERING

Agglomeration Schedule

Stage Cluster Combined Coefficients Stage Cluster First Appears Next Stage

Cluster 1 Cluster 2 Cluster 1 Cluster 2

1 4 5 7.000 0 0 16
2 19 20 14.500 0 0 8

3 2 6 23.000 CLUSTER 10 0 6

4 3 13 33.500 0 0 6

5 1 14 47.500 0 0 9

6 2 3 62.000 3 4 11

7 8 16 78.500 0 0 15

8 11 19 95.667 0 2 13

9 1 15 114.333 5 CLUSTER
0 2 18

10 7 17 133.333 0 0 15

11 2 12 153.033 6 0 12

12 2 9 173.667 11 0 14

13 10 11 195.500 0 8 18
CLUSTER
14 2 18 219.667 3 12 0 16

15 7 8 250.417 10 7 17

16 2 4 284.750 14 1 17

17 2 7 337.013 16 15 19

18 1 10 392.132 9 13 19
CLUSTER
19 1 2 466.050 4 18 17 0
 1

INTERPRETATION

FROM THE ABOVE TABLE IT IS OBSERVED THAT CO-EFFICIENT COLUMN

INDICATE THAT CO-EFFICIENT VALUE (SHORTER THE DISTANCE CLUB INTO
SAME CLUSTER, LONGER THE DISTANCE CLUB IT INTO DIFFERENT
CLUSTER) INFERED THAT NO OF CLUSTERS CAN BE FORMULATED IS 4.

K- MEANS (NON- HIERARCHICAL)

OUTPUT

Cluster Membership
Case Number Cluster Distance

1 1 3.143
2 1 2.735
3 4 2.610
4 4 2.969
5 4 2.883
6 2 3.274
7 3 4.724
8 2 3.659
9 1 4.656
10 3 4.441
11 3 3.784
12 4 4.220
13 2 3.064
14 1 3.779
15 1 4.276
16 2 4.552
17 2 4.905
18 2 5.233
19 3 2.706
20 3 3.274

CLUSTER NOS MEMBERS

HIP
C1 5 1,2,9,14,15
C2 6 6,8,13,16,1
7,18
C3 5 7,10,11,19,
20
C4 4 3,4,5,12
 Final Cluster Centers

Cluster

1 2 3 4

I prefer email to letters 1.6 3.5 2.8 2.8

2.4 3.3 2.2 2.5
I feel that quality comes at a price.

3.6 2.2 3.2 3.8

I think twice before I buy anything.

3.0 3.3 2.6 3.0

T.V is a major source of entertainment for
me & my family

3.6 3.8 2.6 2.5

A car is a necessity, not a luxury

I prefer fast food and ready to use 4.4 3.5 3.4 4.0
products.
2.2 4.0 1.4 3.5
People are more health conscious today
than they were in the earlier generation.

2.4 1.8 4.6 2.8

Entry of foreign companies has increased
the efficiency of Indian Companies

3.2 2.0 1.8 4.5

Women are active participants in my
purchase decisions.

I believe politicians can play a positive role 2.8 4.0 3.0 3.3
in our lives
3.0 3.5 4.2 4.0
I enjoy watching movies in a theatre

If I get a chance I would like to settle 1.6 3.2 3.6 4.0

abroad
2.0 3.8 2.4 4.0
I always buy branded products

2.0 3.8 2.4 4.0

I frequently go out on weekends

I prefer to pay by credit card rather than by 4.2 2.5 1.8 3.8
cash.
Interpretation
C1 C2
I prefer email to letters I prefer letters over email
People are more health conscious today People are less health conscious today
than they were in the earlier generation. than they were in the earlier generation.
Entry of foreign companies has Entry of foreign companies has
increased the efficiency of Indian increased the efficiency of Indian
Companies Companies
Women are not active participants in my Women are active participants in my
purchase decisions. purchase decisions.
If I get a chance I would like to settle If I get a chance I would not like to settle
abroad abroad
I always buy branded products I always do not buy branded products
I frequently go out on weekends I frequently do not go out on weekends
I do not prefer to pay by credit card I prefer to pay by credit card rather than
rather than by cash. by cash.

C3 C4
I prefer letters over email I prefer letters over email
People are more health conscious today People are less health conscious today
than they were in the earlier generation. than they were in the earlier generation.
Entry of foreign companies has not Entry of foreign companies has
increased the efficiency of Indian increased the efficiency of Indian
Companies Companies
Women are active participants in my Women are not active participants in my
purchase decisions. purchase decisions.
If I get a chance I would not like to settle If I get a chance I would not like to settle
abroad abroad
I always buy branded products I always do not buy branded products
I frequently go out on weekends I frequently do not go out on weekends
I prefer to pay by credit card rather than I do not prefer to pay by credit card
by cash. rather than by cash.
 ANOVA

Cluster Error F Sig.

Mean Square df Mean Square df

I prefer email to letters 3.317 3 1.266 16 2.621 .086

I feel that quality comes at a 1.406 3 1.396 16 1.007 .415
price.
I think twice before I buy 2.739 3 1.599 16 1.713 .205
anything.
T.V is a major source of .489 3 1.158 16 .422 .740
entertainment for me & my
family
A car is a necessity, not a 2.322 3 1.640 16 1.416 .275
luxury
I prefer fast food and ready 1.100 3 1.619 16 .680 .577
to use products.
People are more health 7.400 3 .813 16 9.108 .001
conscious today than they
were in the earlier
generation.
Entry of foreign companies 7.522 3 .874 16 8.607 .001
has increased the efficiency
of Indian Companies
Women are active 7.050 3 .788 16 8.952 .001
participants in my purchase
decisions.
I believe politicians can play 1.550 3 1.472 16 1.053 .396
a positive role in our lives
I enjoy watching movies in a 1.417 3 1.269 16 1.117 .372
theatre
If I get a chance I would like 5.239 3 1.077 16 4.864 .014
to settle abroad
I always buy branded 4.972 3 1.252 16 3.971 .027
products
I frequently go out on 4.972 3 1.252 16 3.971 .027
weekends
I prefer to pay by credit card 6.050 3 .866 16 6.989 .003
rather than by cash.

The F tests should be used only for descriptive purposes because the clusters have been chosen to maximize the
differences among cases in different clusters. The observed significance levels are not corrected for this and thus
cannot be interpreted as tests of the hypothesis that the cluster means are equal.

INTERPRETATION
FROM THE ABOVE ANOVA TABLE IT IS FOUND THAT THERE ARE 8 VARIABLE
WHOSE VALUE IS LESS THAN 0.10 AT 10% LEVEL OF SIGNIFICANCE,
DISTINGUISING VARIABLE AT 90% P<0.10
IT IS INDICATED IN RED
IT SHOWS THAT PERTICULAR VERIABLE IS SIGNIFICANT FOR CLUSTER
ANALYSIS
FROM THE ABOVE TABLE OUT OF 15 VARIABLE 8 VARIABLE ARE SIGNIFICANT
DUE TO (P VALUE LESS THAN (<) 0.10)
THESE 8 SIGNIFICANT STATEMENT OR VARIABLE ARE USEFULL FOR FURTHER
CLUSTER FORMULATION.

Number of Cases in each

Cluster

1 5.000

2 6.000
Cluster
3 5.000

4 4.000
Valid 20.000
Missing .000