Detecting Spatial Heterogeneity and Non-Linear Temporal Trends in Violence Against Women Across Rajasthan
摘要
Rape and attempted rape incidents remain critical concerns both within the state and across the country, requiring detailed spatial and temporal assessment. This study examines district-level patterns across 33 districts from 2015 to 2022 using SaTScan’s spatial variation in temporal trends under a discrete Poisson model. Both linear and quadratic temporal models are applied to distinguish steady trends from non-linear dynamics. The results show consistently higher rape cases than attempted rape across districts, with Jaipur and Alwar exhibiting persistently high incidence counts. Four significant high-trend clusters were identified for both rape and attempt to rape. Bhilwara demonstrated the steepest annual increases in both rape (19.24%) and attempted rape (49.28%). The quadratic SVTT approach reveals additional insights beyond linear trends by identifying districts with relatively low current incidence but accelerating growth, particularly in western Rajasthan. These findings highlight the importance of distinguishing between areas of high incidence and those with rapidly increasing trends. The study demonstrates that incorporating quadratic trend analysis enhances the detection of emerging risk patterns, supporting more targeted and forward-looking policy interventions.