SPSS Tutorial

1
SPSS Tutorial

• AEB 37 / AE 802
• Marketing Research Methods
• Week 8 Mario Mazzocchi

2
Principal components and cluster analysis

• Tutorial outline
• Example of principal components analysis
• Example of cluster analysis
• Work on the assignment

3
Open the tutorial dataset

• http//www.rdg.ac.uk/aes02mm
• Go to the SPSS tutorial page and download
  week7.sav
• Open it in SPSS

4
The week7.sav dataset
5
Principal components analysis basic steps

• Select the variables to perform the analysis
• Set the rule to extract principal components
• Give instruction to save the principal components
  as new variables
• Examine output

6
Analyze /Data reduction
7
Select the variables
Press here
Select here
8
Define extraction method
SelectCorrelation matrix
Extraction technique
Click here first
Extraction rule
9
Save components score
Thick this box
Click here first
10
Run the analysis
11
Output(1)
Communalities
12
Output (2)
Interpretazione componenti
13
Components score saved
14
Cluster analysis basic steps

• Apply Wards methods on the principal components
  score
• Check the agglomeration schedule
• Decide the number of clusters
• Apply the k-means method

15
Analyse / Classify
16
Select the component scores
Select from here
17
Ask for the agglomeration schedule
Thick this box
Click here first
18
Select Wards algorithm
Select method here
Click here first
19
Very important
If you do not change the Icicle values, the
Ward algorithm may take ages!!!
Change these values
Click here first
20
Output Agglomeration schedule
21
Number of clusters
22
The dendrogram
23
Number of clusters

• Number of cases 779
• Step of elbow 774
• __________________________________
• Number of clusters 5

24
Now repeat the analysis

• Choose the k-means technique
• Set 5 as the number of clusters
• Save cluster number for each case
• Run the analysis

25
K-means
26
K-means dialog box
Specify number of clusters
27
Save cluster membership
Click here first
Thick here
28
Final output
29
Cluster membership
30
Representation of first two factors(and cluster
membership)