I was wondering, whether there is a way to have statistical significance test for cluster agreement. I know that I can use classAgreement() function to get Rand index, which will give me some indication whether the clusters agree or not, but it would be interesting to have a formal test. Thanks.