Re: Cluster analysis on dataset with ordinal and nominal data

From: Art Kendall (Arthur.Kendall_at_verizon.net)
Date: 02/17/05 

Date: Thu, 17 Feb 2005 13:54:14 GMT

The TWOSTEP cluster procedure in SPSS handles variables at different
levels of measurement. It produces a simple clustering, not a
hierarchical clustering.

Details of how to proceed depend on the details such as what questions
you are using the data to answer, any logical sets of variables, etc.

If you are using Likert scales (sum or mean of a set of Likert items),
the results are very unlikely to discrepant from the interval level of
measurement.

If you have Likert items meant to be be part of scales, you might want
to do item analysis (e.g., with RELIABILITY) and see if they do comprise
a scale or scales.

Age is usually considered a ratio level variable. There is an intrinsic
meaning to zero.

Art
Art@DrKendall.org
Social Research Consultants
University Park, MD USA
(301) 864-5570

cybersurferus wrote:
> Hi,
> I am required to perform cluster analysis on a dataset which has
> ordered category (Likert scale) data as well as ordinal (eg age) and
> nominal (eg race) data. In order to perform the analysis, I plan to
> transform the ordinal and likert scale data to a continuous scale (x)
> with the following function :
> x = (r-1)/(N-1) where r = 1, 2,.... N are the ordinal ranks.
> Also, each nominal variable can be transformed to m binary (0,1)
> variables where m is the number of categories of that nominal variable.
> I then plan to use PROC CLUSTER with some sort of distance measure. My
> questions are :
> 1) What distance measure would be best for these type of variables ?
> 2) Is there any alternative method of cluster analysis of this dataset
> containing both nominal and ordinal variables ?
> 3) Does SAS 9 incorporate Latent class analysis models ?
>
>
> Thanks in advance,
> Pat
>

Relevant Pages

  • Re: Clustering categorical data
    ... cluster analysis, ... your "scales" are internally symmetrical in wording, ... objects/cases here are "items", which are measured on ...
    (sci.stat.math)
  • Re: Finding useful functions- part 1
    ... Bear in mind that I actualy write cluster ... > The one thing one soon learns after using cluster analysis practically ... approach to the functional classification of neurons, ... cells can be clustered so that most of the diversity is captured by the ...
    (sci.cognitive)
  • Re: Cluster analysis
    ... I am working as a tec consultant for Statistica ... having two methods of cluster analysis: ... you can go for k-means clustering method ... Cluster Analysis.Both of these tools can be used for cluster analysis. ...
    (sci.stat.consult)
  • Re: clustering
    ... > going from a five cluster step to a 4 cluster step, ... into the subject of "cluster analysis" about a DECADE after I ... ALL existing clustering algorithms! ... In Art' second sentence, it is imperative to drop the word "easily". ...
    (sci.stat.edu)