The aim of the paper is threefold: (1) to demonstrate the rich repertoire of clustering
capabilities of a ROPstat and R-based new and free software, called ROP-R, by illustrating
several analyses with real psychological data; (2) to show how well ROP-R works in
tandem with ROPstat software in complex classification analyses; and (3) to explore
some nontrivial types of parent attachment using the clustering modules of ROP-R.
Four modules of ROP-R are available for performing cluster analyses (CAs), with several
methods (e.g., divisive hierarchical CA, k-medoids CA, k-medians CA, model-based CA)
not found in other user-friendly menu-driven software. In the paper, mother and father
attachment data are used from a study with adolescents (Mirnics et al., 2021) to illustrate
how the ROP-R software can be used to perform various CAs and evaluate the results
using attractive graphs and useful tables. Comparing different clustering methods,
it was found that both standard AHCA and k-means CA could discover a 7-type structure,
which was also verified by the nonstandard k-medians CA. However, the nonstandard
k-medoids CA and MBCA methods were not very effective in identifying a structure with
an acceptable overall homogeneity. Nevertheless, they were able to identify some types
through extremely homogeneous clusters.
