The role of climate change education on individual lifetime carbon emissions

primary reason for their choices, but we do not know if this was also the case with all students. Further, without a more detailed survey, it is difficult to understand whether other factors (e.g., social circumstances) also contributed to these changes. This is one reason we chose to use the California per capita emission reduction as a control group, so that pro-environmental trends seen throughout California could be accounted for. Other considerations. We also acknowledge a number of other uncertainties in the design of this study. Participants were surveyed at least five years after taking the course, and we recognize the limits of human memory may skew some of their responses. There may be students who incorrectly remember aspects of the course, and this may have influenced some of our conclusions. This is in part why we chose to do a focus group to more accurately investi- gate aspects of the course that may have been important. 7.2. Study limitations One limitation in this study is the lack of a control group or a pre-survey. We acknowledge that without such accompanying data, determining the precise relationship between students’ participation in the course and their current attitudes and behaviors is difficult. We did attempt to control for how pro-environmental behaviors in California have become more common over the last decade, but we do not have any data that measured student attitudes or behavior before taking the class. Although further studies should consider the various ways to

measure changes in participant attitudes and behaviors, measuring such changes over many years remains a challenge. Another limitation in the study is the potential for selection bias. Although we attempted to determine whether students self-selected into the course based on their environmental lean- ings or the 3-unit incentive, we do not have independent data to quantify the role that selection bias had on student enrollment. If students did select this course because of their initial interest in environmental stewardship, this could bias the outcomes of the study. Another concern is related to biases in participant responses to survey and focus group questions. We acknowledge that a socially-agreeable response bias with regard to behaviors being attributed to the course may exist in the participant responses to surveys and focus group questions. Although we took measures in our survey design and focus group protocol to reduce such biases, it cannot be ruled out that such self-reporting response biases may be pres- ent and could influence the reliability of the results. Finally, we recognize that among the uncertainties identified in section 7.1, none of them have been adequately quantified. Although some of these uncertainties, such as the reliability of the carbon footprint calculator and the related carbon emissions, probably would not influ- ence the primary outcomes of the study, other uncertainties such as initial attitudes of partici- pating students may have a larger influence on the study results.

As we have generally described, establishing linkages between an educational campaign and long-term behavior can be challenging. Other studies that attempt to establish causal links between education and environmental quality also faced similar challenges (e.g., [5]), and yet the insights gained from such work provide a strong motivation for environmental education and this type of research [7]; [53]. Our work is similar. Despite the limitations we have identi- fied, our analysis provides important insights into understanding the role that well-designed climate change education can play on long-term attitudes and behavior. 8. Conclusions The potential role of education on individual carbon emissions was studied using data from stu- dents who completed an intensive university course on climate change. Students were surveyed The role of climate change education on individual lifetime carbon emissions PLOS ONE | https://doi.org/10.1371/journal.pone.0206266 February 4, 2020 18 / 23