Cluster Analysis (Quantitative Applications in the Social Sciences),   ISBN:9780803923768

Customer Rating: 

Aldenderfer provides a concise introduction to the various types of clustering methods typically used in the social sciences. If you are a researcher, you really should consult a more comprehensive text. But the current book at least offers brevity.

One key finding from the methods presented here is that the clusters are often subjective. That is, by tweaking various parameters in a given clustering method, you can end up with different clusters. More advanced statistical methods might be needed to discern which agglomerations are likely to be valid associations.

Also, the software described is very badly out of date. Inevitable when the book was written in 1984. Today, the latest packages can handle bigger data sets, and often have much friendly user interfaces.

Customer Rating: 

I can imagine this book would have been very useful in 1984 when it was written, but with the myriad advances in statistical software, not to mention the field of cluster analysis itself, it isn't that much use today. If you are looking to understand and perform cluster analyses with something like SPSS, this book is not at all what you need. If you are looking for a very general understanding of cluster analysis as it was 22 years ago then this might be okay, but otherwise you should look elsewhere.
I feel a bit cheated having bought it on the basis of the other reviewer's glowing recommendation. As a colleague of mine says "An idiot learns from his mistakes, a clever man learns from other people's". Learn from this idiot's mistake and don't buy this book.

Customer Rating: 

I have become a big fan of this little green book series. I belong to a very quantitatively oriented in-house think tank of a major West Coast financial service institution. As a very regular MBA, I often wonder what I am doing in such a group. These little green books have bailed me out several times and provided me the understanding on various esoteric advanced statistical methods. Thanks to these books I taught myself Logistic Regression, and Discriminant Analysis.

About two weeks ago, one of our best Russian mathematicians left our group. He had developed an expertise in Cluster Analysis. My boss assigned me to become his successor as a Cluster Analyst so to speak. If it were not for the green book series, I would have been in a state of panic. I quickly ordered the Cluster Analysis book. Studied it. And, now I am on my way to becoming a descent Cluster Analyst.

There are many eccentric features to Cluster Analysis. First, it is much less well grounded in mathematics and statistics than many other data analysis methods. For one thing, it was invented by biologists at first and further developed by many soft scientists of all kinds. The authors reflect this strange background of Cluster Analysis. One of them is a professor in anthropology, and the other a professor in clinical psychology. Somehow, these soft-scientists have a much greater need to classify their data (this is especially true of biologists) than pure mathematicians do. Second, there are tons of different ways to conduct Cluster Analysis. All have their benefits, and some have specific flaws. The authors do an excellent job at explaining and differentiating these different methods.

As usual [of the green book series], the book is very well written, and makes this complex methodology easily accessible. It is an excellent book to teach you Cluster Analysis. I strongly recommend it.

Cluster Analysis is a good thing to know. These days it is popping out everywhere. How do political campaign managers customize their political message through direct mail to specific voting groups? Cluster Analysis. Whether you are aware or not, we are all part of data clusters. How do college recruiters decide on which applicant to spend much recruiting energy? Cluster Analysis. Cluster Analysis is the answer to numerous unexpected questions.