PropertyValue
?:about
?:abstract
  • "SPSS 11.5 and later releases offer a two step clustering method. According to the authors' knowledge the procedure has not been used in the social sciences until now. This situation is surprising: The widely used clustering algorithms, k-means clustering and agglomerative hierarchical techniques, suffer from well known problems, whereas SPSS TwoStep clustering promises to solve at least some of these problems. In particular, mixed type attributes can be handled and the number of clusters is automatically determined. These properties are promising. Therefore, SPSS TwoStep clustering is evaluated in this paper by a simulation study. Summarizing the results of the simulations, SPSS TwoStep performs well if all variables are continuous. The results are less satisfactory, if the variables are of mixed type. One reason for this unsatisfactory finding is the fact that differences in categorical variables are given a higher weight than differences in continuous variables. Different combinations of the categorical variables can dominate the results. In addition, SPSS TwoStep clustering is not able to detect correctly models with no cluster solutions. Latent class models show a better performance. They are able to detect models with no underlying cluster structure, they result more frequently in correct decisions and in less biased estimators." (author's abstract) (xsd:string)
?:contributor
?:dateModified
  • 2004 (xsd:gyear)
?:datePublished
  • 2004 (xsd:gyear)
?:duplicate
?:hasFulltext
  • true (xsd:boolean)
is ?:hasPart of
?:inLanguage
  • en (xsd:string)
?:linksURN
?:location
is ?:mainEntity of
?:name
  • SPSS TwoStep Cluster - a first evaluation (xsd:string)
?:provider
?:publicationType
  • Arbeitspapier (xsd:string)
?:reference
?:sourceInfo
  • GESIS-SSOAR (xsd:string)
rdf:type
?:url
?:urn
  • urn:nbn:de:0168-ssoar-327153 ()
?:volumeNumber
  • 2004-2 (xsd:string)