Classifying perpetrators of stalking-related behavior with latent class analysis

Abstract

Background

Stalking often escalates from mild behaviors, such as surveillance, to severe forms involving violence, posing serious risks to victims. While identifying high-risk perpetrator classes is crucial for preventing victimization, research on perpetrator classification based on stalking behavior remains limited.

Purpose

This study used latent class analysis (LCA) to classify perpetrators by gender based on their stalking-related behavior (SRB) patterns from a Japanese population survey, and examined the risk levels and characteristics of each class.

Methods

Participants included 530 Japanese individuals (240 men and 290 women) aged 20–49 years whose non-marital romantic partner had expressed a desire for separation and no contact.

Results

LCA revealed four-class models for both men and women. For men, the classes identified were indirect seeker (71.9 %), direct pursuit (21.3 %), serious perpetrator (3.8 %), and mild perpetrator (3.0 %). For women, the classes were indirect seeker (68.2 %), direct pursuit (19.2 %), malicious defamation (10.5 %), and serious perpetrator (2.2 %). In both genders, serious perpetrators exhibited a higher likelihood of violence and suicide attempts, emphasizing the need for mental health assessments and appropriate care alongside violence prevention.

Conclusions

Future research should focus on exploring the diversity of high-risk perpetrators.