The relationship between the natural environment and human society is changing
significantly, with consumption growing hyperbolically. While technological innovation has
helped alleviate some environmental problems, it was insufficient to mitigate them efficiently.
Understanding the factors influencing human behavior is imperative for developing effective
intervention policies to reduce environmental impacts. A promising approach is to design
strategies to endure behavioral changes through collective actions. This research investigated
the influence of social norms on waste prevention behavior at different levels of
environmental motivation through a series of computational experiments using a populationsized sample. Waste prevention behavior (WPB) was analyzed using the General Ecological
Behavior (GEB) scale through a survey assessment. Using real-world data, agent-based
modelling and simulation (ABMS) was conducted to infer the influence of social norms on WPB
in a large-scale experiment. The combination of these methods, alongside a qualitative
approach to understanding the situational factors, provided a comprehensive diagnosis of
individuals’ engagement in waste prevention activities related to avoiding plastic bags in
supermarkets. In the baseline model, 34% of events resulted in agents opting for reusable
bags due to the influence of descriptive norms influenced by their environmental motivation.
The intervention experiment indicated a 22.7% overall reduction in forgetting the reusable
bag compared to the control experiment. The most environmentally motivated groups have
shown significant decreases compared to the other motivation levels. The adoption of
reusable bags increased by 11.2%, with a 4.3% reduction in plastic bags in the intervention
experiment compared to the control group. These results show that intervention policies
combining social norms and levels of environmental motivation can effectively increase
individuals’ engagement and mitigate environmental impacts related to these behaviors.
Moreover, the methodology adopted in this study indicates that computational experiments
using real-world data can overcome some obstacles of large-scale actual experiments by
providing significant details on how behavioral change interventions may influence
engagement levels. Future studies on WPB can use these findings to build real intervention
experiments with more precision to evaluate other possible influences on individuals’
engagement performance.
