[2P1-OS-23] 社会科学・人文科学分野の行動インサイトを活用した機械学習と最適化

[2P1-OS-23-04] Optimization of Financial Incentive Policy for Exercise Adherence

〇Hideaki Kim¹, Takeshi Kurashima¹ (1. NTT)

[[Online]]

Keywords:Financial Incentive, Exercise Adherence, Self-Efficacy

With the wide spread of points programs, financial incentive has been a popular intervention tool to motivating individuals to engage in health behaviors such as physical activity. Traditional strategies of financial incentive have been proposed to offer constant, decreasing, and increasing rewards over time for a daily or weekly behavior goal. But they could be too simple, and much remains to be explored to find more sophisticated incentive strategies for exercise adherence. In this paper, by focusing on a psychological variable called self-efficacy, we investigate cost-effective incentive strategies for maintaining a recommended level of physical activity. We analyze a field experiment data of walking program, and propose a probabilistic model that mimics a participant's behavior in the walking program. Through a simulation study with the proposed model, we show some cases where more cost-effective incentive strategy than conventional approaches could be found.