Exploring Types of Parent Attachment via the Clustering Modules of a New Free Statistical Software, ROP-R.

Area of Science:

• Psychology
• Computational Statistics

Background:

• Parental attachment is crucial for adolescent development.
• Existing software for cluster analysis (CA) lacks user-friendly interfaces and diverse methods.
• The ROPstat and R-based ROP-R software offers advanced clustering capabilities.

Purpose of the Study:

• To showcase the clustering capabilities of the ROP-R software using real psychological data.
• To evaluate ROP-R's performance in complex classification tasks, especially with ROPstat.
• To explore parent attachment types using ROP-R's clustering modules.

Main Methods:

• Utilized four ROP-R modules for cluster analysis (CA).
• Applied various CA methods including divisive hierarchical CA, k-medoids CA, k-medians CA, and model-based CA.
• Analyzed adolescent mother and father attachment data from Mirnics et al. (2021).

Main Results:

• Divisive hierarchical CA and k-means CA identified a 7-type structure, validated by k-medians CA.
• k-medoids CA and model-based CA (MBCA) showed limitations in identifying overall homogeneous structures.
• Nonstandard methods successfully identified highly homogeneous clusters, revealing specific attachment types.

Conclusions:

• ROP-R provides a versatile and user-friendly platform for psychological data analysis.
• The software effectively performs complex classification and explores nuanced attachment patterns.
• Different clustering methods yield varying results, highlighting the importance of method selection in attachment research.