Recap of Session Dated 11th Feb 2012

In todays session sir talked about:
1.Factor Analysis
2.Cluster Analysis
3.Multidimensional Scaling

1.Factor Analysis 

• It is a Positioning technique
• It is Data reduction technique
• It is the clubbing of attribute into meaningfull factor
• Factor is done on variable level

Taking example of Tooth Paste:
      Different attributes of a tooth paste can be : whiteness, cavity, bad smell, freshness, tooth decay, shininess,   strong teeth etc
These attributes can be clubbed as:

1. Asthetic : whiteness,freshness, shininess
2. Health & Hygine : cavity, bad smell, tooth decay, strong teeth

Steps in Factor Analysis

1. Identifying number of factors : This is done with the help of Eigen Value ( Thumb rule for this is 80% of variance)                                                                                                                                      * Eigen Value : It is the amount of variance explained by the attributes
2. Assigning variables to factors : This is done with the help of Corelation Matrix                                     If r > 0.5 , then we can corelate the attributes
3. Naming the factors : Name the factor so that it can be suited to all variables ( It is marketing interpretation)

2. Cluster Analysis 

• It is a Segmentation tool
• Cluster is done on respondent level
• Cluster Analysis is homogenity within the sample and heterogenity across the sample

Steps in Cluster Analysis

1. Identifying number of clusters : It is done using Eucledian distance ( i.e. distance between two clusters). As soon as Eucledian distance increases by great value, then we say that heterogenity between two clusters have been identified.

        
 
     2.  Assigning respondents to clusters : This is done with the help of Dendogram.

    3. Profiling (Market Interpretation)

3. MDS

• Perception Map (XLSTAT)

 Please add any missing points

Thanks & Regards:
Shruti Aryani
Roll No - 80
PGDIE- 41