Assessing and correcting belief about energy sources using reinforcement learning

Addressing the climate crisis requires active citizen engagement and democratic cooperation, rather than top-down impositions by governments. For the public to make rational decisions on pressing issues such as energy policy, they must form and maintain accurate beliefs about key factors influencing these decisions. The goal of this study was threefold: first, to assess the accuracy of public beliefs regarding critical variables—such as energy production, CO2 emissions, and mortality—associated with different energy sources; second, to test the efficacy of a reinforcement-based intervention in correcting potential misconceptions; third to assess whether or not our intervention influenced participants’ preferences. Across seven countries (the UK, US, France, Spain, Italy, Germany, and Japan), we found that participants  held systematic misconceptions about these key variables. However, our reinforcement-based intervention was effective in correcting these misbeliefs. Additionally, we observed that changes in beliefs were accompanied by shifts in preferences, illustrating a pathway from knowledge to action. Follow-up experiments conducted later revealed that these effects persisted over time. Our study suggests that reinforcement-based interventions may be a simple and effective approach to correcting beliefs about critical environmental factors.