Hierarchical Cluster Analysis in SPSS (SPSS Tutorial Video #29) - Dendrogram

In this video I describe how to conduct and interpret the results of a Hierarchical Cluster Analysis in SPSS. I especially emphasize using Ward's method to create more evenly sized clusters, and spend a lot of time on how to interpret a dendrogram with the goal being to identify a reasonable number of clusters to include in a final clustering solution.

This SPSS tutorial series is designed to teach you the basics of how to analyze and interpret the results of data using SPSS. I will cover everything from the very basics of the main windows within SPSS, to manipulating data, to running and interpreting meaningful analyses like t-tests, ANOVA, regression, and many more, and visualizing results.

Follow me at:

Equipment Used for Filming:

[radosawpuakowski]

Excellent tutorial. In two first minutes you have answered my questions and solved my problem. My compliments on your didactic skills.

[danasfar]

This video is SO clear! As short as possible with everything I need to know. THANKS a lot!

[herbhupfer]

Really good job of explaining a lot of options. You spoke VERY clearly which is so important in understanding the nuances of these procedures.

[pietroserrano]

Thank u sir, Im from Portugal and your videos helped me to a Data Analysis test. You're such a good teacher!!!

[dsavkay]

Thank you for the explanation!! Amazing

[sigurd]

Hello Jeff! Thanks for a great video. I'm currently trying to interpret the results of a dendrogram. I'm using a data set provided by my professor, who informed us there should be 7 clusters for our K-means clustering. However, I wanted to check if that would match a dendogram following your instructions, but it doesn't seem to add up. Despite the lack of information do you have an idea about what might cause the difference?

[missioneconomics]

Can cluster analysis be applied on secondary data

[tinageorge]

This is a wonderful video! thank you. During discussion with a statistician they asked me what the cluster analysis fit was? How do i defend (with numbers or a stat) the number of clusters i chose?

[kristingibson]

This is such a helpful video! I am curious as the why you used individual items instead of creating a scale from the items for the cluster analysis? Thanks in advance for your help!

[marie-eveclement]

Thank you for the videos! Very helpfull! Can I use binary variable to enter into cluster analysis?

[the_tons]

Is thee ward method available in the standard spss package?
What version of spss was used in this video? TY❤

[mehdimajidi]

How to remove the gray background from the dendrogram? I want to use the dendrogram for my article... Is there a way to export like only the cluster lines?

[heynadia]

should we include the dependent variable in the analysis?

[Charlatan_]

Great job, thanks so much for these videos. I have question tho. Do you have some author/s so I can justify how I pick my clusters? Thanks so much in advance!

[ehteshamali]

Hello sir. I watched the video. I still do not understand how did you deduce that we have 2 clusters from the dendrogram.

[asdadadadsasda]

Dear Prof., thanks for all the fantastic videos. I would like to perform an cluster analysis with binary data (0 or 1). What algorithm would you recommend for a hierarchical analysis. And you make a video how to perform the silhouette method in SPSS. Thank you very much.

[farshadjahangiri]

Hi there if I want to cluster 30 station with 7 paramateres and time series for 9 years what can I do ? can you help me . this will bi 3 dimensional matrix

[muhammadqasim-bmoj]

Dear Researcher, kindly guide, how can i cluster the questionnaire line items of the large data set. like more than1000 observation?
I have 79 final line items of questionnaire now i want to cluster the line items into distinct latent variable. kindly guide me how can i cluster the line items. thanks in anticipation.

[jeremysayer]

Thank you, your video is very informative, and definitely deserves more views. I was wondering about the clustering of variables together — not on a case-by-case basis, but on a variable-by-variable basis. I assume that, in SPSS, this would give you an indication of the overall clustering of similar variables together, and how similar the data points in one variable are, relative to the others. Is that right? So in your example, it would show groups of similar dimensions of "importance", and their hierarchical linkage to the other dimensions of "importance"?

[horanghae]

Hello, sir! I am currently working on my thesis and this tutorial helps a lot! May I ask, how can I identify the cluster with high similarities of data by inspecting the dendogram? Is it possible or should I apply some calculations to address this? Thank you so much!