Class 3, 4, 5 24/01/10 by Kanishka Singh

The subjects handled in the three classes today were 

1. Hierarchical Cluster Analysis: Cluster Analysis is used to classify cases (e.g. respondents, names or probable customers) into clusters

(groups) based on their variable values. The word variable implies characteristics of these 'cases'. So basically it is a method of analysis which is used to

put similar cases in one cluster so that these clusters can be addressed to or focussed upon.

some of the most simple uses of cluster analysis can be:

a. Marketing segmentation analyses

b. Analyze similarities and differences among new

products

c. Profile resulting clusters for demographic

similarities and differences

It starts with a measure of similarity which is generally the distance between the variables and is completed by deciding on the clustering criteria that is 

variables which are critical to clustering objective.

The next logical step was the result of agglomerative clustering where two or more clusters are added to achieve one agglomerated cluster.

It is opposed to divisive clustering where clusters are broken down.

SPSS was made use of and different clusters were formed using nine sets of variable to result in the formation of a new idea about an offering by using

proximity values and dendrograms.

Perceptual Mapping: The use of Permap software was done to develop perceptual map within variables. Perceptual maps are used to develop a visual image 

of the perceptions of the buyer and collates them to make sense and turn into a real business strategy. First we started off with the example of few soft drinks and their similarities which was determined by a rating process. The resulting data was arranged in a matrix format which was fed as input to the permap software.

It gave a visual understanding of the similarity that soft drinks have as perceived by respondents. Next was attribute perceptual mapping which was certainly more useful as it gave the perception in relation to the attributes and how they contributed to the outcome of another variable.

Thanks and Regards,

Kanishka Singh

Roll No:12140
Operations Batch 09-11