Mean comparison of three clusters for significance

[Joseph Arroyo]

Hello Everyone,

I'm currently conducting a cluster analysis of several countries (cross-section) and would like to compare the mean values of certain variables like average educational attainment and GDP with each other. I have 3 clusters and would like to know if the mean values are significantly different with respect to the clusters, meaning that the GDP Levels are significantly different in cluster 1 and cluster 2. Since I have many variables (arround 20), ANOVA/MANOVA was not very feasible so far.

when I run my code as:

mean values count GDP schooling socialrights rank, over(clusters)

my results dont tell me if they are significantly different, only the std errors and the confidence intervals. Is there a way to add this?


This is my first post here, I'm sorry if I didn't respect any guideline regarding posting.

[Fei Wang]

Joseph, why not using -anova- variable by variable?

If you'd like to implement all the analyses with one line of code, then maybe

mvreg values count GDP schooling socialrights rank = i.clusters

[Joseph Arroyo]

Joseph, why not using -anova- variable by variable?

If you'd like to implement all the analyses with one line of code, then maybe

mvreg values count GDP schooling socialrights rank = i.clusters

Hi Fei,

Thank you for your answer! ANOVA variable by variable would be extremely large and not illustrative for my work.