Hello, My name is Elena and working on cluster analysis for respondents segmentation based on two latent factors. I wonder whether I am doing the right way. I have 400 respondents answering on 2 latent factors, named: Consumer cosmopolitanism (12 items) and Consumer Ethnocentrism (7 items) and all is measured by 5-point Likert scale. How do I perform cluster analysis in this point to group 400 respondents into 2 or 4 clusters?

Currently, I've read from previous articles and analyzed as follow.

• Firstly, compute mean scores for COS scale and CET scale (SPSS)
• Secondly, perform hirarchical cluster analysis with two mean scale scores to determine the optimal numbers of cluster (using Ward's method and Squared Euclidean distance).
• Thirdly, perform K-mean cluster analysis with two mean scale scores and select optimal numbers of cluster (from 2nd step). Check results from Initial Cluster and ANOVA.