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: czcams.com/video/3bAPwFern_4/video.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.
thank you very much. my Master's Thesis is complete coz of you!
Thank you very much, but how did you enter the label in your hierarchical analytical box that presented string results in the dendrogram?
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: czcams.com/video/3bAPwFern_4/video.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)
@@Gaskination How does measurement invariance actually work incase of groups created from the same 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...)
Yes, those are the heterogeneity measures.
@@Gaskination Thank you!
But what that distances mean on the X axis? for example, between the bottom cluster line almost on ten and the right line almost on 25
It represents the distance between clusters. The larger the x-value the larger the distance (difference) between clusters.
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.
@@Gaskination Thanks a bunch, your videos have been a real help
hi , my data set is of 900 observations in which i am not get the line graph , can you help
If you are checking the box for Agglomeration schedule, and it is still not producing the line graph, then I'm not sure. Sorry about that.
is there a way to determine variable importance in hierarchical clustering, like you can in two-step?
flamenkomuva unfortunately, not that I'm aware of.
ok, thanks anyway! and thanks for your videos! :}
shall the grilled chicken nugget separate at the level "3" but not the "4"?
The other way. They separate at level four, but not three.
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.