I have constructed this dendrogram using SPSS. Now I need to explain it properly.
That means, two clusters. Right sir? Ahmed Dheyab Ahmed
Hello Shoaibe,
It looks as if you are attempting to cluster variables/measures rather than the more usual object of cases/samples. Is that intentional?
SPSS-generated dendrograms are not very good at portraying "high definition" with respect to changes in intra-cluster noise, especially at the early end (left-hand side) of the agglomeration history. I'd prefer to look at the clustering coefficients for each stage from start to finish, especially with such a small set of cluster objects.
What the figure suggests to me is, after the third stage, the noise of further agglomerations jumps up pretty dramatically. After that, you're on your own!
But, do remember that: (a) choice to standardize or not; (b) metric chosen for dissimilarity; and (c) agglomeration method all make a difference in the results.
Good luck with your work.
Dear Mose,
I am learning how to do cluster analysis. I am at the very beginning stage. So I need an expert's help to grow to expertize.
Thank you very much for your nice comment and suggestions. Which software do you suggest for making dendrograms?
Hello Shoaibe,
The horizontal axis of the dendrogram represents the distance (dissimilarity) between clusters. The vertical axis represents the objects and clusters. A dendrogram is usually simple to interpret. The horizontal position of each split gives the distance between the two clusters. Looking at this dendrogram, you can see the two clusters as two branches that occur at about the same horizontal distance (~23). You may want to use several distance metric and/or linkage agglomerative methods to understand better the structure of your data .
good luck
E
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS