A Serious Game to Improve the Verbal Resilience Against Doorstep Scams

The primary objective of this initiative is to enhance the verbal resilience of elderly individuals against doorstep scams through the application of serious gaming technology. By providing a secure, repeatable, and accessible environment, the platform enables users to practice assertive communication techniques when confronted with fraudulent actors. The training utilizes a tablet-based architecture featuring six expert-validated, branching scenarios that simulate real-world pressure, allowing users to develop the confidence necessary to refuse suspicious requests and report potential criminal activity.

Technical implementation centers on a sophisticated voice analysis module that employs a support vector machine algorithm to classify user assertiveness with 86.56% accuracy. This system dynamically adjusts scenario outcomes based on both the content of the user's verbal responses and their vocal delivery, creating a responsive feedback loop. While initial evaluations confirmed the high educational value and realism of the scenarios, subsequent iterative testing revealed that while additional gamification elements like reward systems and mascots increased user motivation, they did not necessarily improve core learning outcomes.

The research demonstrates the feasibility of integrating automated voice analysis into safety training for older demographics. Currently deployed by the KBO-PCOB, the application serves as a practical supplement to traditional role-playing exercises. Although the project successfully establishes a framework for training against various forms of fraud, including phone and street-based scams, the authors acknowledge that quantifying long-term behavioral changes in real-world settings remains a significant methodological and ethical challenge. Ultimately, the work highlights the importance of balancing complex feedback mechanisms with a seamless user experience to ensure that training remains both effective and accessible for the target population.