Institute of Management, Christ University
Kengeri Campus,
Bangalore.

Review
On
Cluster Analysis
Submitted for the partial fulfillment of process related to
Continuous Internal Assessment 1
In
Marketing Analytics
[MBA - III Trimester (MBA 322)]

Submitted By
P.Akhilesh 1420419

Cluster analysis?
The Cluster Analysis is an explorative analysis that tries to identify structures within the data. Cluster analysis is also called segmentation analysis .It is used to identify groups of cases if the alignment is not previously known. Because it is explorative it does make any distinction between dependent and independent variables. The different cluster analysis methods that SPSS offers can handle binary, nominal, ordinal, and scale (interval or ratio) data. Different types of cluster analysis include: * Hierarchical Analysis * Two step Cluster * K-means cluster
From the given data set: * We have both continuous and categorical variables * Continuous variables include: Attitude towards Nike, Awareness towards Nike, Preference for Nike Loyalty for Nike and purchase intention for Nike * Categorical variables: Sex,
Hierarchical Cluster Analysis:
Ward method:

* Here it is clearly mentioned that from cluster one 15 respondents are available , from cluster two 15 respondents are available and from cluster three 10 respondents are available * We have 5 missing respondents who had not respondent or the data is inaccurate

Interpretation for Agglomeration table

* The results start with the results start with an agglomeration schedule which provides a solution for every possible number of clusters from 1 to 40 in our example. * The column to focus on is the central one which has the heading ‘coefficients’. Reading from the bottom upwards, it shows that for one cluster we have an agglomeration coefficient of 604.400, for two clusters 413.607, for three clusters 273.533, etc. * If we rewrite the coefficients as in (not provided on SPSS) it is easier to see the changes in the coefficients as the number of clusters increase * The final column, headed ‘Change’, enables us to determine the optimum number of clusters.

Reformed agglomeration table No.of clusters | Agglomeration last step | Coefficients this step | Change | | 604.400 | 413.067 | 191.33 | | 413.067 | 273.533 | 139.534 | | 273.533 | 207.633 | 65.9 | | 207.633 | 176.133 | 31.5 | | 176.133 | 146.346 | 29.78 |

The longest distance seems to be here and this can be the dividing line

Interpretation of Dendrogram using ward linkage: * The y-axis is a measure of closeness of either individual data points or clusters * The horizontal axis of the dendrogram represents the distance or dissimilarity between clusters. * Respondents 40,44,9,15,26,28 are equally distinct from the respondent 2 because 2 is in different cluster * There is huge distance between the respondent 20 and respondent 12 and these can be considered as outliers

Two step Clustering Interpretation:
Two step clustering analysis is more of a tool than a single analysis. It identifies the groupings by running pre-clustering first and then by hierarchical methods. Because it uses a quick cluster algorithm upfront, it can handle large data sets that would take a long time to compute with hierarchical cluster methods. In this respect, it combines the best of both approaches. Also two-step clustering can handle scale and ordinal data in the same model.
Interpretation of two step clustering: * From the below output the quality of the cluster is doing fair and about to reach good * For further interpretation look over the cluster sizes where we have two: cluster 1 and cluster 2 where the respondents are divided into two sizes and their ratio is said to be 1

* For further more analysis use predictor importance for more interpretation : * Attitude towards Nike is considered as high important than the others and predictor importance is given as 1 * Coming to the purchase intention of Nike the predictor importance is 0.5 * * For further analysis : * From below snap we can see that the mean of cluster 1 parameters is higher than the mean of cluster 2 parameters

* The frequently occurring category in cluster 1 is male and in cluster 2 is female