Effects of Solutions Centered Climate Education on Youth Beliefs and Behaviors: The University of California’s Bending the Curve Course

Abstract

Per the United Nations, enhancing climate literacy can play an essential role in advancing climate mitigation, adaptation, and promoting sustainable human behaviors. Yet, there is a lack of empirical research explicitly studying the effects of climate solutions focused education. Here, we studied the effects of a climate solutions focused course—the University of California Bending the Curve (BtC) course on: (1) climate change beliefs, (2) personal pro-environmental actions, and (3) psychological health, using baseline and post-course surveys. A total of 374 youth (median age 21 ± 1.7 years, 63% female) participated in the study, and data analysis focused on statistically comparing pre- versus post-course survey-based data. We observed that the BtC course enhanced climate change beliefs. Specifically, at post-relative to pre-course, we observed significantly increased belief that global warming will impact individuals personally as well as impact our future generations; it tripled the number of students who believe that humans can and will act to reduce global warming; it significantly increased the number of individuals who believe in a scientific basis for climate change. Notably, climate solutions education also enhanced belief in the efficacy of personal climate action and increased agreement amongst youth that many of their friends also share the same views on global warming. With regard to personal pro-environmental actions, the course significantly improved self-reported actions, including waste reduction, making food choices with reduced emissions, and purchase of carbon offsets. These actions reduced the carbon footprint per student at post- vs. pre-course by a significant 0.3 ± 0.1 CO₂ tons/year, which is equivalent to the CO₂ absorbed by about 15 trees per year. While psychological health outcomes did not show any significant post- vs. pre-course change, we found that enhanced personal pro-environmental actions as well as enhanced psychological health were predicted by course-related strengthening of climate change beliefs. Overall, our findings provide evidence that solutions-based climate education can be an important strategy to enhance individual climate change awareness as well as personal pro-environmental actions that lead to significant individual carbon footprint reduction, with potential for widespread scale-up.

1. Introduction

Anthropogenic climate change refers to the unsustainable fossil-fuel-dependent human growth practices that have accelerated global warming and heightened the frequency and intensity of weather extremes and natural disasters such as wildfires, floods, and sea level rise [1]. Over the last four decades, the planet’s temperature has increased by over 0.5 °C and will continue to rise with continued emissions, leading to worsening risks to health, livelihoods, food and water security, and economic costs [1]. Studies have documented the negative consequences of natural disasters and climate change-related weather events not just on human physical health but also on mental health [2,3,4,5]. People affected by climate disasters suffer from higher rates of anxiety, depression, post-traumatic stress disorder, stress, sleep disruption, substance abuse, and even suicidal behaviors [6,7,8,9,10,11,12,13]. Furthermore, these physical and mental impacts can persist months to years after a climate disaster event [12]. Given accelerated warming and frequent climate disasters, there is a ‘chronic fear of environmental doom’ in the general population, defined as eco-anxiety, which is particularly pronounced in youth [5,14]. Indeed, a recent survey of 10,000 youth across 10 countries confirmed that climate change is associated with climate anxiety and distress in youth globally and is threatening their health and well-being [15]. Given this scenario, there is an urgent need to build climate resilience, including mitigation approaches, which rapidly curb emissions and reduce future risks, as well as adaptation approaches that protect us in the face of unavoidable climate risks [16].

The United Nations has identified climate education for the public as an essential element for building awareness and support for climate resilience globally [17]. As a result, climate change education courses have expanded to focus on both the science behind global warming and climate change solutions [18,19,20]. Climate-specific education has been shown to foster positive attitudes towards the environment [21] and enhance belief in the existence, causes, and potential impacts of climate change [22]. Climate education can further increase individual willingness to adopt pro-environmental behaviors, such as recycling, reducing energy consumption, and support for climate-friendly policies [23]. Moreover, individuals who perceive a greater impact of climate change on their lives, in terms of health, safety, and wealth, engage in more pro-environmental behaviors, such as adopting more sustainable living standards, making climate-friendly food choices, paying green taxes, and purchasing higher-priced ecofriendly products [24,25,26,27]. Greater environmental concern driven by climate change has also been linked to increased intention for urban forest conservation [28] and willingness to adopt sustainable agricultural practices [29]. In this context, persuading individual perceptions that the world is changeable via our actions as opposed to fixed has been shown to enhance response efficacy towards pro-environmental behaviors [30].

Yet, there is a dearth of quantitative evidence specifically showing the effects of climate solutions focused education on individual beliefs and behaviors [31,32]. Few prior climate education studies have been small pilots evaluating course feasibility or have focused on post-course evaluations but without baseline data, which leaves unanswered how climate education, especially with a solutions focus, shapes belief and behavior outcomes relative to prior knowledge [33,34,35]. It also has not been answered whether emphasizing solutions focused learning can empower the mental health of learners, which may be crucial in the current times of widespread climate distress [36]. Hence, here, we conducted baseline as well as post-course survey-based evaluations for a climate solutions focused course, the University of California Bending the Curve (BtC) course, specifically surveying climate change beliefs, pro-environmental decision-making, and psychological health.

Theoretical framework. The theoretical underpinnings for this study stem from many sources. These include Climate Change Communication and Education Theory, which emphasizes the role of climate literacy in shaping public understanding and engagement [37]. Psychological theories also apply, including theories that propose the influence of attitudes on behavior, such as the Theory of Planned Behavior [38]—in this case environmental attitudes shaping pro-environmental behavior—and Value-Belief-Norm theory [39], which emphasizes that individuals’ values, beliefs, and personal norms influence their pro-environmental behaviors. Further, the Stimulus-Organism-Response framework [40] also generally applies, as it explains that external stimuli, such as environmental messaging, influence an individual’s behavior by affecting their internal state, i.e., awareness and attitudes. From a pedagogic perspective, Constructivist Learning Theory [41], which suggests that learners build understanding through active engagement and reflection, also forms the basis of this work. Ultimately, while several theories support this investigation, we did not design this study to generate support for any one of these specific theories. Here, we aimed to empirically evaluate how climate solutions education may stimulate sustainable behaviors, reduce carbon footprint, and improve personal well-being and resilience.

2. Materials and Methods

The study was conducted over two years (2022–2024) on 374 students enrolled in the BtC course at the University of California San Diego (UCSD) (n = 370) and a handful of students enrolled at the University of California Riverside (UCR) (n = 4). The study sampling method was to survey all BtC enrolled students at two time points: once at baseline (i.e., pre-course) and once at the end of the 10-week course lasting one academic quarter (i.e., post-course). The study’s target population was thus students enrolled in the 10-week BtC course, and the time interval between the pre- and post-course surveys was 10 weeks, which has been shown to be a reasonable time period for new habit formation and initiating behavior change [42]. The UCSD Institutional Review Board (IRB) committee approved this human subject research, and all participants consented to the survey.

The sample size for this study was adequately powered to detect small effect size differences in pre- vs. post-course outcomes (Cohen’s d > 0.15) statistically compared between two dependent means (matched pairs) at a two-sided alpha level of 0.05 and 0.8 power, calculated using G*Power v3.1 software [43].

2.1. University Course

Participation in the study was offered to every University of California student who took Bending the Curve (BtC)—a climate solutions focused course created in 2018 [19,44]. This interdisciplinary course was offered across departments such that students who took the course ranged in their backgrounds from engineering, natural sciences, and marine biology to humanities and social sciences, including political science. The course is delivered as an upper-division course, i.e., for college students beyond their first two years of education, suggesting a level of prior knowledge or specialization, but not necessarily dedicated within the environmental/climate sciences.

BtC focuses on climate change solutions to “bend the warming curve” and accelerate resilience and climate justice for our planet’s most vulnerable people. Climate resilience is the ability of an individual, family, organization, and/or community to withstand, respond to, recover from, and learn from the impacts of climate change [35]. The course introduces ten scalable solutions that emphasize carbon neutrality and climate stability (see

Table 1. Ten scalable solutions integral to the Bending the Curve course that emphasize carbon neutrality and climate stability.

Solutions	Key Actions
1. Reduce short-lived climate pollutants (SLCPs) and replace fossil-fuel energy systems with carbon-neutral technologies	Immediate reduction of SLCPs; sustainable replacement of fossil fuels with carbon-neutral systems.
2. Foster a global culture of climate action	Coordinate public communication and education from local to global scales.
3. Deepen global climate collaboration	Design venues where stakeholders, community, and religious leaders converge to collaborate with multidisciplinary academic scholars to mitigate climate disruption.
4. Scale up subnational governance and collaboration models	Replicate successful subnational climate initiatives to embolden and energize national and international action.
5. Adopt market-based instruments	Create efficient incentives for businesses and individuals to reduce CO2 emissions; implement cap-and-trade or carbon pricing; use cost-containment mechanisms; apply direct regulation where market-based structures do not credibly exist.
6. Narrowly target direct regulatory measures at high-emission sectors	Use rebates, efficiency, and renewable portfolio standards; remove harmful subsidies; build political coalitions in favor of climate policy; expand support for low-emission innovation in high-emission sectors not covered by market-based policies.
7. Promote widespread use of mature low carbon technologies	Deploy photovoltaics and wind turbines, promote EVs, and promote more efficient end-use devices in lighting, air conditioning, appliances, and industrial processes.
8. Support innovations for full electrification and efficiency	Aggressive support and promotion of advanced low-cost energy storage, smart grids, and microgrids; accelerate the complete electrification of energy and transportation systems; improve building efficiency to target 80% CO2 reduction by 2050.
9. Reduce methane and black carbon emissions	Cut methane by 50% and black carbon by 90% through maximum use of available technology and regulation; achieve co-benefits of climate and health
10. Regenerate damaged ecosystems and restore soil organic carbon	Proactive afforestation, reduce deforestation, improve natural sinks for organic carbon soil, implement food waste reduction programs, and use energy recovery systems to maximize utilization of food produced and recover energy from unconsumed food.

) and fall under six categories: science, societal transformation, governance, economics, technology, and ecosystem management. Together, these foci create an integral approach to designing climate change solutions across multiple disciplines. Notably, global deployment of these practical measures has the potential to reduce 20 percent of the current 50 billion tons of emissions of CO₂ and other greenhouse gases and, in addition, meet the United Nations sustainable development goals by creating wealth for the poorest 3 billion on planet earth.

Practically, the BtC interdisciplinary curriculum consists of a large multimedia content library co-created by a team of researchers, educators, and designers at the University of California, and led by UC San Diego climate scientist Veerabhadran Ramanathan, UC San Diego political theorist, Fonna Forman and online education innovator, based at UC Online, Scott Friese. The course uses a flipped classroom teaching style where students watch pre-recorded multimedia lectures at home, recorded by dozens of researchers across disciplines, with class time dedicated to activities and discussions facilitated by the course professors and teaching assistants [45]. BtC was designed this way to enhance engagement and deepen understanding via active discussion and to support individual learning paces.

The course is taught over one 10–15 week-long quarter/semester. It is taught by thirty-nine professors and experts across the UC campuses, with expertise pertaining to different aspects of climate change solutions. The full curriculum is divided into 47 modular units, each featuring a studio-produced video lecture, assignments and in-class activities, quizzes, supplemental readings, and instructor resources. Resources also include an open-access digital textbook with dedicated chapters for each of the modules [19].

The curriculum was initially launched across five UC campuses and is now offered on seven of the nine undergraduate campuses, and more than 2000 UC students have completed the course. Further, the course is available as a massive online open course (MOOC) and has been licensed by six other universities and other institutions that prioritize climate education for their members and students, including the American Medical Association [46], the One Health Workforce Academies [47], and most recently, the Los Angeles Unified School District.

Overall, aspects of the course that distinguish it from other climate education include its main focus on interdisciplinary and scalable climate solutions rather than solely on the science of climate change. Its multifaceted approach considers diverse aspects of climate change, including ethical and intergenerational equity issues, highlighting the interconnectedness of the problem and the need for various stakeholders to work together. BtC emphasizes a “solutions-thinking and design” framework across fields of science, economics, governance, and social behavior, and the curriculum equips students with the knowledge and skills to actively participate in addressing climate change. It empowers students to become active change makers, encouraging them to design and implement their own solutions via active project-based learning, which in turn promotes both self- and collective efficacy.

2.2. Surveys

A set of surveys was delivered at two time points—pre (within the first week of the start of the course) and post (during the last week of the 10-week course). Self-report surveys were delivered using the online HIPAA-compliant Research Electronic Data Capture (REDCap) quantitative survey system; we did not have the opportunity or resources to observe actual behaviors exclusive from the self-reports, which is a limitation of the study. The surveys were divided into three sections, assessing (1) climate change beliefs, (2) personal pro-environmental actions, and (3) psychological health.

(1) Climate change beliefs (CCBs). CCB was based on a published assessment [33]. It queried knowledge, beliefs, and attitudes about climate change and was specifically adopted as it has been previously used in the context of studying the impact of climate education on students. (2) Personal pro-environmental actions (PA). This survey was also based on a published assessment [33] and probed students’ individual actions to mitigate climate change. This survey was used because it quantitatively links survey responses to individual carbon footprints, hence providing an important objective measure for climate action. Personal actions include (a) reduce-reuse-recycle practices to check the amount of personal waste produced; (b) practices to reduce energy consumption at home, such as use of renewable energy and use of energy-efficient lighting; (c) dietary choices to reduce carbon emissions; (d) transportation choices with reduced emissions; and (e) purchase of carbon offsets for flight travel. Depending on the question type, responses were made either as multiple choice selections (questions 1, 2, and 4) or on a 5-point Likert scale (questions 3 and 5).

Quantitative personal action data from the PA survey was used to calculate the approximate carbon footprint reduction relative to emissions made by an average person residing in California. Carbon footprint reduction was calculated for each individual at each time point (pre- vs. post-), per details provided in Appendix A 

Table A1. Description of the assumptions and reductions in carbon emissions for each of the personal pro-environmental actions survey responses.

Survey Response	CO2 Reduction (Tons/Year)	Assumptions Used in Cool Climate Calculator
Buy renewable energy	1.36	100% electricity from renewables
Buy energy star appliances	0.04	Energy star fridge or other products
Install solar panels	0.68	50% electricity from renewables
Install solar hot water	0.4	50% of heating of water from solar hot water
Change light bulbs	0.16	5 bulbs used 5 h per day
Buy hybrid/electric car	4.22	average of hybrids (40 mpg × 2/3) + electric vehicles (99 mpg × 1/3)) = 59.67 mpg
Recycle more often, buy products with less packaging, compost food scraps, give away/donate products	0.42	Reduce waste by 25% for each action
Carpool regularly	0.85	3 times/week
Buy more fuel-efficient vehicle	2.08	32 mpg vs. 22 mpg for 13,100 miles/yr.
Use public transit more	0.42	20 miles/week in bus instead of 22 mpg in car
Use bike for transportation	0.53	20 miles/week instead of 22 mpg in car
Make food choices to reduce emissions	0.69	Response of ‘all the time’ or ‘often’ is default setting for ‘low carbon version of American diet’ (0.69 tons/year). Response of ‘sometimes’ or ‘occasionally’ is a reduction of 0.46 tons/year through increase in meat consumption from 244 calories (default) to 353 calories.
Buy carbon offsets for flying	0.93	Response of ‘all the time’ or ‘often’ is 80% of flights purchased offsets (0.93 tons/yr.). Response of ‘sometimes’ or ‘occasionally’ is 40% of flights purchased offsets (0.46 tons/yr.).

[33]. Specifically, each personal action was correlated with a certain carbon footprint reduction from the norm that was used to calculate overall carbon footprint reduction at pre- and post-time points. Percentage carbon footprint reduction was also calculated for each personal action at both pre- and post-time points. This percentage was assessed as the relative contribution of each action to the overall carbon footprint reduction for each student. The percentage reduction attributable to each action was calculated such that the combined total across all five actions summed to 100% for each student.

(3) Psychological health. Three standard surveys were used to probe psychological health; specifically, the Short Warwick-Edinburgh Mental Well-being Scale (SWEMWS) [48] was used to assess well-being, the Perceived Stress Scale (PSS) was used to measure psychological stress over the past month [49], and the Brief Resilience Scale (BRS) was used to assess participants’ ability to bounce back from stressors, i.e., their stress resilience [50]. These surveys were used as they have been previously shown to have sound psychometric properties, i.e., are valid and reliable measures. Further, outcomes of climate-related stress, well-being, and resilience are all relevant to a psychological and mental health framework for climate resilience [51]. Additionally, we assessed reliability for these Likert scale-based survey measures at baseline using the Cronbach’s alpha measure of internal consistency, which showed high reliability (Cronbach’s alpha > 0.8) for all three instruments (SWEMWS: 0.80, PSS: 0.86, BRS: 0.81). As the Likert scale range varied for individual CCB questions and PA was based on checkbox responses, Cronbach’s alpha was not calculated for these scales.

2.3. Data Analyses

Data were analyzed using JASP, MATLAB v2025a, Jupyter notebook, and Google Sheets. Pre- vs. post-course survey data were not normally distributed; hence, they were summarized as the median and median absolute deviation (MAD) of responses across all subjects. This is because median rather than mean values provide a more robust measure of central tendency for nonparametric data. Percent responses for each response category were also calculated. Pre- vs. post-course data were statistically compared using the Wilcoxon signed rank test, which is appropriate for comparing related samples of nonparametric data. False discovery rate (fdr) corrections were applied for multiple comparisons. Pre- vs. post-course comparisons were performed, and as we predicted, course-driven enhancement of climate change beliefs, greater personal pro-environmental actions, and potentially improved psychological health were observed at post-course per the theories mentioned in the Introduction section.

We also used linear regression models to investigate whether pre- vs. post-course changes in climate change beliefs predict course-related change in pro-environmental actions and psychological health. Here, we predicted that change in climate change beliefs would strengthen pro-environmental action, and potentially such change in attitudes would also enhance psychological health. F-statistic, significance of fit, and significant model coefficients were reported.

3. Results

3.1. Demographics

Students’ demographic data are shown in

Table 2. Demographics of youth participating in the Bending the Curve climate solutions course during 2023–2025. MAD: median absolute deviation.

Demographics	N = 374
Age (years, median (mad))	21 (1.7)
Gender n (%)
Male	140 (37)
Female	234 (63)
Ethnicity n (%)
Caucasian	136 (36)
Black/African American	7 (2)
Asian	136 (36)
Native Hawaiian or Other Pacific Islander	1 (0)
American Indian/Alaska Native	6 (2)
More than one ethnicity	41 (11)
Other	47 (13)

. A total of 374 students completed both pre- and post-surveys.

3.2. Climate Change Beliefs

A summary of results of the CCB survey is provided in

Table 3. Climate change beliefs (CCBs) across 374 course participants. CCB question statements and responses to anchors are shown as percentages (%) of total responses, alongside the pre- and post-median and median absolute deviation (MAD) of responses. Wilcoxon signed rank p-values (p-val) comparing pre vs. post responses are shown and are fdr-corrected for multiple comparisons. Response anchors discussed in the results text are highlighted in grey.

Statement	Not at all (1)/Only a Little (2)		A Moderate Amount (3)		A Great Deal (4)		Do Not Know (5)		Pre Median (MAD)	Post Median (MAD)	p-value
	Pre %	Post %	Pre %	Post %	Pre %	Post %	Pre %	Post %
CCB1: How much do you think global warming will harm future generations?	1.9	0.5	8.0	5.6	89.8	93.3	0.3	0.5	4 (0.2)	4 (0.1)	0.013
CCB2: How much do you think global warming will harm you personally?	12.3	8.3	53.7	47.9	33.2	42.8	0.8	1.1	3 (0.6)	3 (0.6)	<0.001
Statement	1/2		3		4		5		Pre Median (MAD)	Post Median (MAD)	p-value
	Pre %	Post %	Pre %	Post %	Pre %	Post %	Pre %	Post %
CCB3: Which of the following statements comes closest to your view?	1.3	0.8	28.9	13.6	64.4	69.8	5.4	15.8	4 (0.5)	4 (0.3)	<0.0001
CCB3 key: 1. Global warming is not happening. 2. Humans cannot reduce global warming, even if it is happening. 3. Humans could reduce global warming, but people are not willing to change their behavior, so we are not going to. 4. Humans could reduce global warming, but it is unclear at this point whether we will do what is needed. 5. Humans can reduce global warming, and we are going to do so successfully.
Statement	Strongly agree (1)/Somewhat agree (2)		Neutral (3)		Somewhat disagree (4)/Strongly disagree (5)				Pre Median (MAD)	Post Median (MAD)	p-value
	Pre %	Post %	Pre %	Post %	Pre %		Post %
CCB4: The actions of a single individual will not make any difference in global warming.	24.9	17.9	17.1	12.0	58.0		70.1		4 (1.0)	4 (0.9)	<0.0001
CCB5: New technologies can solve global warming without individuals having to make big changes in their lives.	30.7	33.2	19.0	15.5	50.3		51.3		4 (1.1)	4 (1.2)	n.s.
CCB7: I have personally experienced the effects of global warming.	71.9	76.2	15.0	13.4	13.1		10.4		2 (0.8)	2 (0.7)	n.s.
Statement	1		2		3		4		Pre Median (MAD)	Post Median (MAD)	p-value
	Pre %	Post %	Pre %	Post %	Pre %	Post %	Pre %	Post %
CCB6: Which of the following statements comes closest to your view?	84.8	93.9	1.3	1.1	8.0	4.6	5.9	0.5	1 (0.6)	1 (0.2)	<0.0001
CCB6 key: 1. Most scientists think global warming is happening. 2. Most scientists think global warming is not happening. 3. There is a lot of disagreement among scientists about whether or not global warming is happening. 4. Do not know enough to say.
	None (1)		A Few (2)		Many (3)				Pre Median (MAD)	Post Median (MAD)	p-value
	Pre %	Post %	Pre %	Post %	Pre %		Post %
CCB8: How many of your friends share your views on global warming?	4.3	1.6	25.4	24.6	70.4		73.8		3 (0.5)	3 (0.4)	0.023

. Among the eight CCB questions, six showed significant change in participants’ beliefs, and these are further illustrated in Supplementary Figure S1. Notably, we also attempted to reduce CCB data to a few key components using principal component analysis, but the results suggested that up to five factors were necessary to explain 90% of the data variance; given this large number of components, we chose to report outcomes for each CCB item separately.

Specifically, CCB1 asked students if they believed climate change would impact future generations. We found that 93.3% of students believed climate change would have a great deal of impact (response option 4) after taking the BtC course in comparison with 89.8% who made this choice before taking the course.

CCB2 assessed whether students believed climate change would harm them personally. Based on our data, there was an increase of nearly 10 percentage points in the number of students who believed climate change could impact them a great deal: 33.2% to 42.8% (response option 4).

CCB3 assessed whether students believed global warming was occurring and whether humans could reduce it. We found that the number of students who believed that humans are unwilling to change their behavior decreased by more than half: 28.9% to 13.6% (response option 3). Simultaneously, the number of students who believed that humans were going to successfully reduce global warming tripled: 5.4% to 15.8% (response option 5).

CCB4 assessed whether students agreed with the following statement: “The actions of a single individual will not make any difference in global warming”. Here, we found an increase of over 12 percentage points in students who disagreed with this statement: 58.0% to 70.1% (response options 4 and 5).

CCB6 assessed students’ beliefs regarding scientific consensus about whether global warming is happening. We found an increase in the number of students who believed in the scientific consensus that global warming is occurring: 84.8% to 93.9% (response option 1).

Finally, CCB8 determined if participants’ friends shared their beliefs, and here, we found an increase of three percentage points for the category of many friends with shared beliefs at post-relative to pre-course: 70.4% to 73.8% (response option 3).

The remaining two CCB questions had insignificant changes between pre- and post-course; these were CCB5: New technologies can solve global warming without individuals having to make big changes in their lives, and CCB7: I have personally experienced the effects of global warming.

3.3. Personal Pro-Environmental Actions

The second section of the survey focused on personal pro-environmental actions to reduce carbon emissions (

Table 4. Students’ pre and post responses for personal pro-environmental actions. Response anchors that showed significant change are highlighted in grey. χ² tests were applied to checkbox item questions 1, 2, and 4, and Wilcoxon signed-rank tests were used for questions 3 and 5. p-values are fdr corrected for multiple comparisons. * Other energy savings responses included being mindful of energy consumption, reducing energy use especially in peak hours, switching off lights and appliances in unused rooms, installing auto on/off light sensors, and hang-drying clothes.

(1) Waste Reduction		Pre %	Post	p-value
Recycle more often		90.1	91.7	n.s.
Buy products that have less packaging		52.4	65.5	<0.001
Compost food scraps or food waste		40.9	44.9	n.s.
Donate products so that they can be reused		76.5	74.3
Other		8.8	7.0
No actions taken		1.3	1.1
(2) Energy Savings		Pre %	Post	p-value
Purchased renewable energy from utilities		21.7	24.1	n.s.
Purchased energy-saving appliances		40.4	44.1
Installed solar panels		21.1	19.5
Installed solar hot water		4.8	5.1
Changed traditional light bulbs to energy-efficient light bulbs		54.6	61.0
Other *		3.7	10.7	<0.001
No actions taken		23.5	20.3	n.s.
(3) Food Choices
Never/ Occasionally		Sometimes/Often/ All the Time
Pre %	Post %	Pre %	Post %	p-value
52.1	37.2	47.9	62.8	<0.001
(4) Transportation Emissions		Pre %	Post	p-val
Purchased a hybrid car		18.7	16.8	n.s.
Carpool regularly		45.2	52.7
Purchased a more gas-efficient car		23.8	24.6
Used public transportation more often		58.8	64.2
Used a bicycle instead of a car as transportation		17.9	20.1
No actions taken		16.0	13.4
(5) Carbon offsets
Never/ Occasionally		Sometimes/Often/ All the Time
Pre %	Post %	Pre %	Post %	p-value
88.0	76.7	12.0	23.3	<0.001

). Significant pre- vs. post-course change was observed for 4 of 5 categories, i.e., for waste reduction, energy savings, low carbon emission food choices, and purchasing carbon offsets, but not for transportation emissions.

For waste reduction, buying products with less packaging was significantly reduced by 13 percentage points. In energy savings, BtC course students took many actions not listed in the survey, including being mindful of energy consumption, reducing energy use in peak hours, and switching off lights and appliances, with a significant 7 percentage point reduction in energy use. Regarding food choices, we found that BtC students making low-emission food choices increased by about 15 percentage points. Finally, for carbon offsets, the percentage of BtC students who sometimes, often, or always purchased offsets when flying increased by 11 percentage points. Supplementary Figure S2 further shows plots for the significant changes in personal actions.

Here, we also acknowledge that while we measured personal pro-environmental actions, the extent of such personal action is limited and importantly shaped by systemic factors that we did not measure, such as infrastructure, economic, or social norms that can either facilitate or hinder personal efforts to reduce emissions.

3.4. Carbon Footprint

Calculation of individual carbon footprint reduction from the norm, i.e., relative to an average person in California, was based on the personal pro-environmental action survey data. Three of the five categories of personal action showed significant carbon footprint reduction at post-relative to pre-BtC. Specifically, improvements were observed for waste reduction, food choices, and purchase of carbon offsets (

Table 5. Individual carbon footprint reduction from the norm based on personal actions. Carbon footprint reduction was measured for each student at pre- and post-course, shown in average (standard deviation) units of CO₂ tons/year per student. The percentage of carbon footprint reduction for each personal action relative to the total carbon footprint reduction was also calculated at pre- and post-course for each student and is shown as the average (standard deviation) percentage across all students. The overall carbon footprint reduction percentage is always 100%, hence left blank. Significant pre- vs. post-changes observed for waste reduction, food choices, and carbon offsets are highlighted. p-values are fdr-corrected for multiple comparisons across all five personal action categories. The overall carbon footprint reduction per student at post- vs. pre-course was 0.3 ± 0.1 CO₂ tons/year and was significant.

	Carbon Footprint Reduction per Student CO2 Tons/Year (SD)
	Pre	Post	Pre %	Post %	p-Value
Waste Reduction	1.1 (0.4)	1.2 (0.4)	26.3 (22.5)	26.1 (18.5)	0.0017
Energy Savings	0.6 (0.7)	0.6 (0.7)	13.5 (14.8)	13.4 (15.3)	n.s.
Food Choices	0.4 (0.3)	0.5 (0.2)	8.9 (11.2)	10.1 (10.5)	<0.001
Transport Emissions	2.0 (2.2)	2.0 (2.1)	48.5 (25.1)	46.1 (22.5)	n.s.
Carbon Offsets	0.1 (0.2)	0.2 (0.3)	2.8 (5.8)	4.3 (6.4)	<0.001
Overall	4.2 (2.7)	4.5 (2.6)	-	-	0.0013

).

We additionally calculated the cumulative student carbon footprint pre-BtC and post-BtC, which was 7835.7 CO₂ tons/year vs. 7726.6 CO₂ tons/year, respectively. Thus, we found a cumulative carbon footprint reduction of 112.1 CO₂ tons/year by BtC course students. On a per-student basis, our data demonstrates an average pre-BtC carbon footprint of 21.0 CO₂ tons/year and a post-BtC carbon footprint of 20.7 CO₂ tons/year; this equals an average reduction of 1.4% per student or 0.3 ± 0.1 CO₂ tons/year.

3.5. Behavioral Changes: Stress, Well-Being, and Resilience

The final section of the survey assessed participants’ perceived stress, well-being, and resilience per standard surveys. No significant changes were detected for these measures (

Table 6. Students’ self-reported psychological health outcomes measured at pre- vs. post-course. Median ± median absolute deviation (MAD) of responses are shown. There were no significant pre- vs. post-changes in these measures.

Outcome	Pre (Median ± MAD)	Post (Median ± MAD)	p-Val
Perceived Stress	20 ± 4.84	20 ± 5.00	n.s.
Well-being	3.43 ± 0.46	3.29 ± 0.45
Resilience	3.33 ± 0.55	3.33 ± 0.55

).

3.6. Relationship Between Outcome Measures

Specifically, we were interested in investigating whether pre- vs. post-course changes in climate change beliefs predict course-related changes in personal pro-environmental actions and psychological health. Linear regression models showed that for pro-environmental actions, enhanced climate change beliefs at post relative to pre-course predicted overall improvements in waste reduction (F-stat = 4.5, p < 0.0001), energy savings (F-stat = 2.6, p = 0.01), food choices (F-stat = 5.1, p < 0.0001), and carbon offsets (F-stat = 3.1, p = 0.002); transport emissions were not affected by the course and were not found to be related to CCB items. Specifically, waste reduction was predicted by CCB1 (global warming will harm future generations, ß = 0.40 ± 0.17 (standard error), p = 0.02), CCB2 (global warming will harm personally, ß = −0.22 ± 0.09, p = 0.02), and CCB4 (actions of a single individual will make a difference in global warming, ß = −0.22 ± 0.09, p = 0.02). Energy savings was predicted only by CCB7 (personal experience with the effects of global warming, ß = 0.18 ± 0.06, p = 0.004). Food choices were predicted by CCB3 (humans can reduce global warming, ß = 0.24 ± 0.12, p = 0.04), CCB4 (actions of a single individual will make a difference in global warming, ß = 0.17 ± 0.06, p = 0.002), and CCB7 (personal experience with the effects of global warming, ß = 0.20 ± 0.07, p = 0.002). In addition, carbon offsets were only predicted by CCB6 (most scientists think global warming is happening, ß = −0.21 ± 0.08, p = 0.008). Thus, different CCB items affected distinct aspects of personal action.

While psychological health did not show overall significant change, enhanced climate change beliefs at post relative to pre-course also predicted alleviated stress (F-stat = 3.5, p = 0.0007) and improved well-being (F-stat = 2.6, p = 0.009) but not resilience. Specifically, stress reduction was predicted negatively by CCB2 (global warming will harm personally, ß = −1.78 ± 0.52, p = 0.0007) and positively by CCB3 (humans can reduce global warming, ß = 1.58 ± 0.56, p = 0.005). Improved well-being was significantly predicted only by CCB3 (humans can reduce global warming, ß = 1.09 ± 0.36, p = 0.003). Course-related change in personal pro-environmental actions did not relate to changes in psychological health.

4. Discussion

This study focused on quantifying the immediate impacts of a solutions focused climate education course on youth climate change beliefs and personal pro-environmental actions and further assessed impacts on psychological health. We found that the course significantly positively affected student beliefs and pro-environmental actions and even led to significant reductions in individual carbon footprint, but with no significant change observed for psychological health outcomes.

According to the first climate change beliefs survey, after taking the course, 93.3% of students believed climate change would greatly affect future generations. The proportion of students who believed that climate change will personally affect them also increased from 33.2% to 42.8%, suggesting greater awareness of climate impacts. Further, belief in the scientific basis for global warming increased from 84.8% to 93.9%. Notably, the solutions focused course tripled the number of students who believe that humans are going to successfully reduce global warming (5.4% to 15.8%) and more than halved the number of youth who were skeptical that humans can reduce global warming (28.9% to 13.6%). Additionally, the course significantly increased the proportion of youth who believed in individual actions by 12%. In addition, participant responses showed a small but significant social impact in that the proportion of students who said their friends also agreed with their views increased by 3.4%. These findings support prior research specific to climate education showing that climate literacy enhances belief in the existence, causes, and potential impacts of climate change [22]. These findings also align with research demonstrating that individuals can be persuaded to understand the world as changeable via our actions as opposed to fixed, and that such beliefs of a changeable world foster intention to act on climate change [30].

Per the survey on personal pro-environmental actions, the proportion of students who made more environmentally conscious food choices increased by ~15%. There was a 13% increase in students who contributed to waste reduction, particularly by buying products with less packaging. Energy savings increased by 7%, for which specific student responses included being mindful of energy consumption, reducing energy use especially in peak hours, switching off lights and appliances in unused rooms, installing auto on/off light sensors, and hang-drying clothes. Finally, a significant 11% change was also observed for the purchase of carbon offsets when flying. These results support evidence that climate education increases individual willingness to adopt pro-environmental behaviors [23], and that when individuals are aware of the impacts of climate change on themselves and future generations, they are more likely to adopt sustainable living standards, especially making climate-friendly food choices, for which we observed the largest impact [24,25,26,27].

We translated the student-reported personal actions into the reduction of carbon footprint made by each student relative to an average Californian and found that students significantly reduced their carbon footprint post-course relative to pre-course by an average of 0.3 CO₂ tons/year. This reduction in emissions per person is considered to have a moderate impact and is equivalent to CO₂ absorbed by about 15 trees per year [52,53]. The University of California education system currently has over 295,000 students enrolled across their campuses [54]. As a hypothetical, if every one of these students were to take the Bending the Curve course, then at 0.3 CO₂ tons/year reduction in emissions, there would be an overall CO₂ reduction of over 88,500 tons per year, which is substantial and equivalent to removing annual CO₂ emissions from approximately 20,000 US households or 400,000 miles of petrol-powered driving. Based on CO₂ equivalents calculations, this is also roughly equivalent to curbing the annual CO₂ emissions from a medium-sized US city [55]. Such scale-up calculations emphasize the importance of climate solutions based education yet must be taken with caution given individual variance and unclear long term effects of behavior change.

The last section of our survey focused on psychological health, specifically probing perceived stress, well-being, and resilience. This is particularly important to assess in current times when the climate crisis is associated with widespread distress, especially among our youth [5,14]. We did not find any significant change in these measures. For such change, dedicated psychosocial eco-resilience curricula may be explicitly needed to improve psychological well-being in the context of climate change [56]. To address this unmet need, the senior author recently co-led the design and implementation of the “Climate Resilience” course focusing on mental health resilience offered across all University of California campuses, which has shown several robust benefits for psychological well-being [36]. This education required dedicated contemplative pedagogy curricula in mindfulness, empathy, compassion, active listening, nature connectedness, and other important inner resilience skills taught by expert mindfulness teachers. Built on the science of personal and social resilience and climate action, students in this course were also required to engage in a collective action project in groups, fostering a sense of belonging and motivating empowerment and collective self-efficacy. From this, we can infer that different climate education course materials may benefit different aspects of personal action vs. personal resilience. Notably, in regression analyses we found that strengthened climate change beliefs at post-relative to pre-course predicted enhanced pro-environmental actions as well as stress alleviation and well-being. Specifically, greater awareness of climate change-driven harm to self and future generations and a strengthened belief that humans can reduce global warming and that actions of single individuals matter predicted BtC course-related change in pro-environmental actions and psychological health.

Overall, the Bending the Curve course was implemented in California, a state that already leads in climate change and sustainability policies in the USA [57], yet it demonstrated that continuing education on climate solutions can lead to meaningful and measurable outcomes. Per the course educators, the project-based focus of the course that refracts substantive content through the lens of student interest, motivation, and passion for solutions is key to the BtC course impact. Students are not passive recipients of information but active climate solutions designers in the classroom, which increases agency and personal investment and may account in large measure for the observed impacts. Since the study was conducted in a region with relatively high acceptance of climate change, our findings raise important questions about how such education might impact regions with greater climate skepticism. As the course is now available as a massive online open course (MOOC), future research may explore its effects in different cultural/socio-political contexts. A version of this course has now also been approved by the Los Angeles Unified School District (LAUSD), the second largest school district in the United States with over 500,000 K-12 students. It will soon be part of a content library of climate-related educational resource materials for specific LAUSD high school educators called “Climate Literacy Champions” to develop age-appropriate classroom education. Hence, this study provides initial data that as climate change impacts more people, solutions focused education may be an important and cost-effective method to change individual beliefs and behaviors about bending the emissions curve.

Prior research has indeed shown that environmental education and enhanced climate change knowledge increase environmental responsibility that further mediates interest in carbon neutrality [21,58]. Research also shows that climate education that includes ethical considerations such as climate justice and inequity issues promotes critical thinking skills to resist climate-related misinformation and manipulation and stimulates deeper involvement in climate and sustainability actions [59]. Our climate solutions centered education study extends this evidence base with a significantly larger sample size than this prior research.

Limitations and future research directions. Our study is certainly not void of limitations. First, it was conducted exclusively among college-aged students, of median age 21 years (range 18–44 years), and primarily at 1 of the 10 University of California campuses, i.e., in San Diego. This relatively homogenous sample limits the generalizability of results to other cities, states, regions, or individuals of differing ages, cultural and political identities. Future research should replicate this work in alternate settings, including non-college populations and institutions across different states and counties, to explore whether similar effects emerge.

Second, the course was optional, raising the possibility of selection bias. Due to the voluntary nature of enrollment, students who took the course may have had previous interest in the subject matter, which could make them more open to behavior change and climate action than students with less prior awareness. This suggests that our study outcomes may not be representative of students who would not choose to enroll in such a course. Yet, of note, at the University of California San Diego, where this course was primarily implemented, there is a requirement for all undergraduates to engage in climate education [60], and selection bias due to self-interest may be less at play in this setting.

Third, while we can reasonably believe that the personal pro-environmental action results and reduction in CO₂ emissions were driven by the BtC course given that the surveys were taken by students in the first week (pre) and last week (post) of class, without a control group we cannot completely rule out spontaneous change or other external factors. While acquiring control group data were not feasible within the scope and funding for this study, future research designs should be based on causal logic, incorporating appropriate comparison groups, and, where possible, objective behavioral and physiological measures. Ideally, such studies would compare BtC outcomes to alternate forms of climate education, particularly those with similar student-engaged, flipped classroom designs.

Fourth, the study, similar to much prior research in this domain, relied on self-reported measures of pro-environmental actions and of psychological health. This reliance on self-reported data introduces potential bias, as students may have over- or under-reported their beliefs and behaviors given the context of participating in a climate-focused course and certain outcomes having socially desirable elements. While a comparison control group can mitigate some of these concerns, it would also be useful to consider objective tracking of behaviors using smartphone apps and wearables, as we and others have implemented in prior research [34,61,62].

Finally, it has been shown that without broader systems changes, i.e., changes in the physical infrastructure, political landscape, and economic systems, it is unclear whether changes in climate beliefs and behaviors will be sustained [20,51,63,64]. One study, however, found that pro-environmental actions were significant even five years after a climate education course [33]. In this regard, it is a limitation of our study that we could not measure systemic influences or long term effects of behavior change, making this an important recommendation for future research. That said, the feasibility of long term measurements needs to be carefully considered in initial course planning, as many students stop responding to assessments after they graduate or simply do not feel incentivized to respond to surveys from previous classes.

In a recent perspective [51], we emphasize that alongside bottom-up climate literacy, we need systemic change brought about by greater connectivity between involved sectors and intentional coalition-building among key stakeholders aiming towards long term policy-making. The BtC curriculum directly addresses this challenge by focusing on interdisciplinary solutions that leverage systemic and structural change, including change in economic systems, environmental policy, and governance. The course additionally emphasizes “societal transformation” as a crucial aspect of bending the curve, implying a recognition that large-scale changes in behavior, attitudes, and institutions are essential. It further explores policy and market mechanisms to incentivize technological innovation and scale clean technologies globally and also considers climate justice as important for addressing inequities that are often rooted in systemic issues. Course discussions often focus on addressing the perceived “false choice between individual behavior change and systems change”, asserting that both are essential for effective climate action. Students are encouraged to see personal responsibility and agency as well as be creative in designing solutions, while being educated on the complex interactions between individual actions and broader societal systems and structures that shape how climate change will be addressed. Ultimately, meaningful progress in addressing climate change requires a combination of individual actions and systemic change. Social structures, real-world incentives, and the inherent characteristics of environmental resources alongside information, habits, attitudes, norms, and influence how we behave and set the limits of behavior change [65]. Our outcomes did not capture these complex effects that may emerge in the long term, but future research should examine how individual behavior shifts led by courses such as BtC intersect with wider societal transformations.

Theoretical Implications. This research contributes to established theoretical frameworks while also highlighting their limitations within a complex system-based context. First, it corroborates Climate Change Communication and Education Theory [37] that emphasizes the role of climate literacy in shaping understanding of climate issues and engagement in climate action. It further supports theories of behavior change such as the Theory of Planned Behavior that fundamentally proposes that attitudes drive intentions and actions [38]. That the course influences climate change beliefs that then impact pro-environmental behaviors is aligned with the Value-Belief-Norm theory [39] and generally also the Stimulus-Organism-Response framework [40]. The current research extends these theories by demonstrating how solutions-based education can serve as the stimulus to significantly increase perceived behavioral control and proactive response that is quantifiable by the average 0.3 CO₂ tons/year emissions reduction achieved per student. Yet, it highlights the limits of such individual action and the urgent need for simultaneous systemic change to achieve global climate goals.

Pedagogically, our study also contributes to Constructivist Learning Theory [41], which supports active project-based learning instead of passive knowledge-building, by demonstrating course efficacy within the specific context of climate change. The findings suggest that the BtC pedagogical model, with flipped-classroom project-based learning and a focus on interdisciplinary solutions, offers a powerful mechanism for students to actively construct their own knowledge. The positive changes observed in participants’ pro-environmental beliefs and behaviors highlight a key theoretical implication—that the most effective climate education is not merely about conveying facts, but about fostering a solutions-oriented mindset that empowers individuals to act. This study therefore provides a framework supporting the constructivist premise that deep, meaningful learning is an active process that translates into real-world application.

Our study also extends the importance of eco-anxiety and psychological health theories [66,67]. These theories also emphasize solution-focused education and that such education may buffer against climate distress by fostering efficacy and agency. Yet we did not find any pre-to-post course change in psychological health outcomes. This null finding underscores the importance of actively integrating mental well-being curricula within climate change education to alleviate eco-anxiety and distress [36,51].

Policy Implications. The findings from this study have significant implications for future climate education policy. Our research supports recommendations for global policy initiatives to require integration of climate change solutions focused education into all climate mitigation efforts. Specifically, policymakers should prioritize the integration of active student-engaged, project-based, and interdisciplinary pedagogical approaches at all educational levels, as these provide an evidence-based mechanism for translating climate knowledge into measurable pro-environmental action. A key policy recommendation is to mandate climate education initiatives that are designed to measure outcomes beyond knowledge acquisition. Thus, policymakers should require the evaluation of how these programs affect attitudes, beliefs, and quantifiable pro-environmental behaviors within the larger societal system.

Another critical area for refinement in climate education policy would be to tackle the psychological burden of climate change by administering classroom programs that are equipped to alleviate climate distress. These educational programs should directly address individual well-being through supplemental and psychological support materials. For climate education to be truly effective and sustainable, it must holistically equip youth with both the knowledge to act and the resilience to cope with climate-related distress.

The United Nations greening education initiative aligns with certain suggestions made here and focuses on integrating environmental sustainability into education systems worldwide [68]. Indeed, some countries, such as Italy and New Zealand, have mandated climate change education in schools, and countries such as Mexico, Argentina, and the U.K. have made important strides as well [69]. In the US, climate change education is mandated in school curricula in the states of New Jersey and Oregon [70,71]; New Jersey specifically requires that climate change be integrated across all K-12 subjects, not just science. These policy efforts are applauded and must be scaled alongside objective evaluation studies to assess the impacts of such climate education on both individual and collective action and global climate resilience.

5. Conclusions

The Bending the Curve course outcomes provide encouraging data and insights into how climate education, particularly solutions centered education, has the possibility to be a successful bottom-up strategy for carbon emission reductions. We acknowledge the several limitations of this study and emphasize thoughtful, larger-scale evaluation. With careful implementation and tailoring for regional/cultural relevance and integrating curricula to address important psychological health concerns, such education could have wide-ranging impacts on climate change mitigation.