Improvement of attitudes and skills using a MOOC about the basic science of climate change

The study addresses how education can mitigate and adapt to the climate crisis by developing Climate Change Competence (C3) among educators and the public. Despite growing awareness of climate change, many people hold partial or poorly formed understandings, which can impede action. The authors propose C3—a framework aligned with EU key competences and similar to GreenComp—comprising knowledge, abilities, and attitudes to enable informed decisions and actions on climate mitigation and adaptation. The paper investigates whether a MOOC focused on basic climate science can improve all C3 dimensions and whether improvements differ by socio-demographic characteristics (teacher vs non-teacher, Spain vs Latin America, age). The importance of robust assessment is emphasized, using a validated Likert-scale instrument to measure C3 evolution and correlations between dimensions.

The paper reviews MOOCs as scalable, open tools for Climate Change Education (CCE), noting roughly 50 CC MOOCs released in the past decade. Prior studies indicate MOOCs can increase knowledge, awareness, climate literacy, and stimulate interdisciplinary learning and discussion, though completion rates are often low and outreach to learners with low prior knowledge can be limited. Socio-demographic factors influence CC awareness and action: gender, age, and regional contexts shape perceptions and pro-environmental behavior; Latin Americans often perceive CC as a present, local issue. This context informs expectations for differential C3 improvement and underscores the need to tailor educational interventions to diverse audiences.

Design: Pre-experimental design without control group, with independent-group pre- and post-tests (anonymized responses) to assess MOOC impact on C3. Mann-Whitney U tests compared pre vs post; Kruskal-Wallis with Dunn’s post hoc (Bonferroni/Holm) compared multiple groups. Effect sizes: rank-biserial correlation (r) for MWU; eta-squared from H for Kruskal-Wallis; Cohen’s d reported for pairwise age comparisons. Sample: Second edition of the Spanish-language MOOC (March 2019) “Awareness and capacity building for Climate Change for Primary and Secondary school teachers.” Pre-test n=530; post-test n=255. Demographics (pre-test): Latin America 58.9%, Spain 38.1% (others ~3%); professions: teachers 46.5%, students 26.4%, non-teaching professionals 27.1%; gender: 51.3% female, 48.7% male; age 20–77 (Mean 36.1, SD 13.6); 74.7% university educated. MOOC: Six modules (10 h each) covering climate system basics, evidence, physical-chemical mechanisms (energy balance, greenhouse effect, carbon cycle), human causes (sectors, energy, food), future scenarios (models and projections), and educational mitigation/adaptation actions. Short videos, interactive forums, paced but self-directed progression; Creative Commons resources. Instrument (C3Q): 36 items across three dimensions with 1–4 Likert scales; higher scores indicate higher C3. - Knowledge (14 items; response options: False to True): subdimensions—Biophysical Processes (BPP), Causes (CAU), Consequences (CSQ), Adaptation & Mitigation (A&M). Items based on climate literacy studies. - Ability (9 items; Never to Always): subdimensions—Purchases (PUR), Transport (TRA), Energy Savings (ESA); frequency of mitigation-related behaviors. - Attitude (13 items; None to Much): subdimensions—Trust per agent (TRU), Responsibility (RES), Educational Support (EDU). Reliability and validity: McDonald’s omega for attitude = 0.81; confirmatory factor analysis for attitude three-factor model adequate fit (χ2=73.95, df=62, p<0.142) with loadings TRU r=0.55–0.71, RES r=0.55–0.90, EDU r=0.66–0.78. Entire scale fit adequate (χ2=1150.44, df=581, p<0.001) with r values: knowledge 0.71; ability 0.68; attitude 0.87; overall omega=0.87. Data analysis performed with SPSS 26; significance at 5%.

Overall C3 improvement: All three dimensions (knowledge, ability, attitude) significantly improved post-MOOC with small-to-moderate effect sizes. - Dimensions (Table 3): - Knowledge: Pre M=3.24 (SD 0.41, n=515), Post M=3.48 (SD 0.40, n=258); U=87295.0, Z=-7.49, p<0.001, r=-0.34. - Ability: Pre M=3.16 (SD 0.49, n=530), Post M=3.37 (SD 0.52, n=258); U=84244.5, Z=-5.31, p<0.001, r=-0.24. - Attitude: Pre M=3.62 (SD 0.39, n=530), Post M=3.73 (SD 0.32, n=258); U=80127.5, Z=-3.98, p<0.001, r=-0.17. Subdimensions (Table 2): - Knowledge: Largest gains in Biophysical Processes (Med 3.00 to 4.00; r=-0.34, p<0.001). Causes improved least but significantly (r=-0.10, p=0.021). Consequences (r=-0.12, p=0.007) and A&M (r=-0.25, p<0.001) improved. - Ability: Purchases (r=0.41, p<0.001) and Transport (r=0.37, p<0.001) showed moderate improvements; Energy Saving small improvement (r=0.17, p=0.017). - Attitude: Trust (r=0.24, p<0.001) and Educational Support (r=0.20, p=0.009) improved; Responsibility not significant (p=0.075; r=0.23, near threshold). Sociodemographic comparisons: - Teachers vs Non-teachers (Table 4): Teachers started lower in all dimensions but improved across knowledge (Pre M=3.22; Post M=3.27) and also showed gains in ability and attitude; non-teachers showed significant improvement mainly in knowledge; post-test differences between groups reduced in ability and attitude (post ability MWU p=0.177; post attitude p=0.214). - Spain vs Latin America (Table 5): Latin-American participants had higher pre-test medians in ability (Z=-4.17, p<0.001, small r=-0.22) and attitude (Z=-3.83, p<0.001, r=-0.20). Both regions improved in all dimensions; post-test differences significant only for ability, favoring Latin America (Z=-5.44, p<0.001, moderate r=-0.40). Knowledge improvements similar across regions. - Age and Ability (Table 6): Kruskal-Wallis H(3)=9.65, p=0.022; median ability by age: 20–34=3.22, 35–49=3.11, 50–64=3.00, ≥65=3.22. Dunn-Holm pairwise: 20–34 vs 35–49 (p=0.040, d=0.20) and 20–34 vs 50–64 (p=0.025, d=0.26) showed small effects; ≥65 not significantly different from 20–34. Indicates higher ability scores in youngest and oldest groups relative to mid-aged groups.

The MOOC, though focused on foundational climate science, led to significant gains not only in knowledge but also in ability and attitude, suggesting interdependence among C3 dimensions. Improved understanding may foster emotional engagement and pro-environmental behaviors, aligning with prior findings in CCE. Notably, educational support attitudes increased, supporting the integration of CC content into formal curricula. Teachers benefited substantially, likely due to professional familiarity with MOOCs and the course’s interdisciplinary integration addressing common teacher challenges in CC instruction. Regional differences show Latin-American participants’ stronger ability and attitude baselines and larger ability gains, possibly reflecting greater perceived immediacy of CC impacts and curricular emphasis. Age effects indicate higher self-reported mitigation-related abilities among the youngest and oldest groups, potentially linked to contemporary climate movements and earlier waves of environmental activism, respectively.

The study demonstrates that a Spanish-language MOOC on the science of climate change effectively improves Climate Change Competence (C3) across knowledge, ability, and attitude. The findings support including C3 in formal education to catalyze climate action and highlight MOOCs as accessible, scalable tools for teacher training and public education. Teachers showed larger relative gains, and Latin-American participants exhibited higher baseline and post-course ability and attitude scores than Spanish participants. Age differences suggest distinct cohorts’ readiness for mitigation behaviors. Future research should broaden samples, compare knowledge-based with other intervention types, and conduct longitudinal studies to assess sustained behavior change and real-world climate actions arising from increased C3.

Generalizability is limited by the voluntary, self-selected sample and higher attrition typical of MOOCs; the 65+ age group had a small n. Pre- and post-test responses were anonymous and analyzed as independent groups, which constrains causal inference at the individual level. Participants’ prior interest and awareness may bias outcomes. Additional studies with broader, more representative samples, alternative intervention formats, and longitudinal designs are needed.