Statistics and Research Methodology for Managerial Decisions: Cluster Analysis Essay Example

<< <

Steps to use cluster analysis

1. First, choose the sample that requires clustering.
2. Next, clearly define the variables that will be measured. These variables can include objects, events, or entities.
3. Then, calculate the similarities between different entities using techniques like correlation and Euclidean distances.
4. Select mutually exclusive clusters based on these calculated similarities.
5. Finally, compare and validate the formed clusters.

Premises of cluster analysis

In order for cluster analysis to work effectively, it depends on distance measures. It assumes that if variables are on the same dimensions or units, they will have similar means. However, if different dimensions or units are used for measurement or if the variables are not similar in nature, this can impact the resulting clusters. To address this issue, standardization can be applied to equalize the impact of variables measured on different scales.

Bunch processs and methods There are two main processes in bunch analysis. They are

• Hierarchical
• Non- Hierarchical

Hierarchical method is structured like a tree, known as a dendrogram. This method consists of the following:

• Agglomerate

This approach begins with each instance being treated as an individual cluster, and at each stage, similar clusters are merged together. It culminates in a single cluster.

• Dissentious

The text begins with a single instance and then divides the instances based on their differences. It concludes with all the instances separated. The

agglomerate method is further divided into:

1. 
   Linkage methods
   

1. Single linkage

Minimal distance or nearest vicinity

2. Complete linkage

Maximal distances or furthest vicinity

3. Average linkage

Average distances between linkages

2. 
   Centroid methods
   

Distance between two centroids.

3. 
   Ward’s methods
   

Squared distance from the agencies.

Non-hierarchical method ( k-means bunch )

K-means bunch provides more stable bunchs, since ; it is an synergistic process as compared with hierarchal method. It needs a pre-specified figure of get downing points to acquire an initial place ; hence, it is best suited in continuance with hierarchal method.

The process of cluster analysis in computer applications is used to segment applications in the research field. The benefits of the internet include availability of updated information, easy navigation across websites, prompt online ordering and query handling, easily comparing prices and services from multiple sellers, access to competitive and educational information about products/services, increase in speed of gathering information from vendors/suppliers, reduction in ordering cost and processing time, and reduction in paper flow. The SPSS process for cluster analysis involves typing in the desired variables according to the input date and proceeding with hierarchical cluster analysis. This includes classifying through the chink analysis menu and choosing

variables to direct to the variable List Box by clicking the right pointer button.

In a similar manner, follow the same steps for other variables. Select the cluster and show according to your needs. Click on the statistics… button to open its sub-dialog box. Check the Agglomeration Schedule, along with any other desired parameters. Proceed by clicking on Continue to close this sub-dialog box. The previous dialog box will reappear. Then choose the Plots… button to open its sub-dialog box.

Select the desired secret plans and click on "Continue" to close this dialogue box. The previous dialogue box will reappear. Click on the "Method..." button to open its dialogue box for the bomber. Choose the desired method and click on "Continue" to close this dialogue box. The previous dialogue box will reappear. In the "Hierarchical bunch analysis" dialogue box, click on "Oklahoma" to view the resulting viewer in K-Means Cluster Analysis.
To classify "K-Means Cluster...", go to the "Analysis" menu and open the K-Means Cluster Analysis dialogue box.
Choose which variable you want analyzed and move it to the variable list by clicking the right pointer button. Repeat this step for other variables as well.
Specify how many clusters you want.
Click on the "Iterate..." button in order to open its dialogue box for the bomber. Select your preferred number of loops, then click on "Continue" to close this dialogue box.The previous one will be displayed again.
To access options for Statistics, select "Options…". After choosing your required options, click on "Continue" to close this annoying dialogue box.The previous dialogue box will reappear.Click on "Oklahoma" to see the final outcome spectator.

Mention Textbooks:

• "Business research methods" by Donald
  

R Cooper - Pamela S Schindler

-