I don't think I understand the question. Do you mean the 1, 2, 3, etc. in the blue boxes? It's just a powerpoint square shape that I typed in. The labels on the dendrogram are placed their automatically by SPSS. If you don't see those, perhaps this will help: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-3bAPwFern_4.htmlsi=brvFqLZ2g0rZkUnO
Professor, Can I do MGA on the groups created after cluster analysis. And do I need to do measurement Invariance for this kind of groups because they all belong to the same study population. Please reply?
Yes, you can do a multigroup analysis on a model using the cluster membership number as the grouping variable. An invariance test will still confirm the measurement invariance (which is not guaranteed within a single sample, even if they are all from the same study population)
@@chefberrypassionateresearcher Even within the same population there are sub-populations. For example, within a company, there are IT workers and accountants. These are different.
Are the coefficients in the table the same as the heterogeneity measures or how can I determine the individual heterogeneity measures for each cluster solution (Like for 2 clusters, 3 clusters etc...)
Hi James, my data set has around 400 observations and produces quite a smooth line graph from the agglomeration schedule. Do you have any advice as to how to determine the appropriate number of clusters in this case?
In this case, choose a useful number of clusters. 1 or 2 clusters is usually not very helpful. More than six becomes hard to interpret. So, usually 3-5 or six is the best choice.
if you look at the distance (X-axis), the distance between grilled chicken and other two occurred even earlier than the second branch, which means they are not "level 4" separation but should be "level 2" instead. You cannot judge a separation simply based on the balancing/number of nodes.