Smartphone-based Intervention for Repeat Offender Rehabilitation and Recidivism Reduction

Thesis Abstract

This study addresses the persistent challenge of repeat offending and recidivism by examining a smartphone-based intervention designed to support rehabilitation and reduce reoffending among convicted individuals. The central aim is to evaluate whether a mobile platform integrating sustained monitoring, cognitive-behavioral micro-interventions, real-time risk assessment, and secure reporting mechanisms can lower six-month and twelve-month recidivism rates compared to standard supervision. Specific objectives include (1) to assess the feasibility and acceptability of the smartphone-based intervention among participants; (2) to determine the impact on proximal risk factors such as impulsivity, aggression, substance use, and pro-social activity; (3) to examine changes in recidivism outcomes at 6 and 12 months; (4) to identify mediating processes (e.g., engagement, perceived coercion, self-regulation) linking intervention exposure to behavioral change; and (5) to evaluate cost-effectiveness relative to conventional supervision. The study employs a mixed-methods design, combining a quasi-experimental, non-equivalent groups approach with a six-month RCT pilot within a larger cohort study. The population comprises adult offenders under community supervision in metropolitan regions, with a target sample of 420 participants (210 intervention; 210 control) for the quasi-experimental phase and a nested randomly selected sub-sample of 120 participants (60 per arm) for the RCT component. Data collection instruments include a) standardized criminological measures (Self-Reported Delinquency Scale, Substance Use Disorder Identification Test), b) risk assessment tools (Level of Supervision Assessment, Short Form of the Spousal/Peer Conflict Scale adapted for offender contexts), c) engagement analytics from the smartphone platform (daily check-ins, GPS-based compliance, app interaction logs), and d) qualitative interviews (semi-structured) with a purposive sub-sample to explore user experience and perceived effectiveness. Validity and reliability are addressed through triangulation across self-reports, official records, and platform analytics; instrument reliability is established via Cronbach’s alpha and test-retest procedures, and data integrity is ensured through secure, encrypted data handling aligned with ethical standards. Quantitative analyses utilize descriptive statistics, propensity score matching to balance groups on baseline covariates, Cox proportional hazards models for time-to-reoffending, logistic regression for dichotomous recidivism outcomes, and linear mixed-effects models to examine longitudinal changes in proximal risk factors. Mediation analyses apply structural equation modeling to test hypothesized pathways from engagement to changes in risk factors and subsequent recidivism. The qualitative component employs thematic analysis to elucidate user experiences, barriers to sustained use, and perceived behavior change mechanisms, with rigorous coding reliability checks and member checking for credibility. Expected findings include a statistically significant reduction in reoffending rates at 12 months in the intervention group, moderated by higher engagement levels and moderated by factors such as digital literacy and housing stability. Anticipated reductions in proximal risk factors (substance misuse, impulsivity, and aggression) and improvements in pro-social activities are expected to mediate the relationship between app usage and recidivism outcomes. The study contributes to knowledge by integrating theory-driven cognitive-behavioral interventions within a scalable ICT-based platform tailored for offender rehabilitation, thereby advancing evidence on digital supervision, ecological momentary interventions, and data-informed risk management. Theoretically, the research draws on Rational Choice Theory and Self-Regulation Theory to interpret decision-making processes and executive function improvements, while incorporating Deterrence and Social Cognitive Theory to contextualize behavioral change within structured supervision. Practical implications include policy guidance on incorporating smartphone-enabled supports into community supervision frameworks, optimization of engagement strategies, and a preliminary cost-effectiveness assessment to inform resource allocation. The main conclusion envisaged is that smartphone-based interventions can feasibly augment traditional supervision, producing measurable reductions in recidivism through sustained engagement and targeted modification of proximal risk factors. Recommendations emphasize scalable deployment, iterative refinement of digital content to match user needs, integration with existing case-management systems, and further research to identify subgroups most responsive to ICT-enabled rehabilitation.