Chapter 8 SPSS Cluster Analyses Utrecht （第八章SPSS集群分析乌特勒支）.pdf

Chapter 8 Cluster Analysis SPSS - Cluster Analysis Datafile used: vcr.sav. This datafile is about the quality of the 2 1, fictional, brands of VCRs. How to get there: Analyze Classify … Hierarchical Cluster… This procedure attempts to identify relatively homogeneous groups of cases (or variables) based on selected characteristics. For example: cluster television shows into homogeneous groups based on viewer characteristics. In hierarchical clustering, an algorithm is used that starts with each case (or variable) in a separate cluster and combines clusters until only one is left. This menu selection opens the following Hierarchical Clustering Analysis main dialog window. The use of this window will vary substantially depending on whether you choose to Cluster Cases or to Cluster Variables; here we focus on Clustering Cases. To cluster cases you need to identify variables you wish to be considered in creating clusters for the cases. The variables to be used for cluster formation are here: picture quality (5 measures), reception quality (3 measures), audio quality (3 measures), ease of programming (1 measure), number of events (1 measure), number of days for future programming (1 measure), remote control (3 measures), and extras (3 measures). Pass these in the Variable(s) box. Then you need to specify how you wish your cases to be identified. This will usually be an ID number or some identifying name (here the variable brand). See the second picture above. Button Statistics … Here you can define the Cluster membership: None: All clusters are listed since all possible solutions will be identified. What it does not do is identify cases included in each cluster for a particular solution. Single solution: Specify some number greater than 1 that indicates cluster membership for a specific number of clusters. For instance if you type 3