Climate Change Conceptual Change Research Studies

Examining the Stasis Theory of Climate Change Cognition

Stasis theory is the widely held assumption that providing people, particularly Americans, with scientific information about climate change is either futile or counterproductive due to the overwhelming dominance of cultural and political context. Researchers such as Kahan have argued that efforts to promote science literacy cannot be expected to dissipate public conflict over climate change. Ranney and Clark (2016) challenge this view, proposing that culture and scientific knowledge interact synergistically. They argue that stasis theory is historically na€ıve, as evidenced by the rapid acceptance of heliocentrism, the spherical Earth, and the links between tobacco and illness despite initial suppression. Furthermore, while evolution has faced a $150-year$ "fighting retreat," the effects of climate change are becoming salient much faster, and the risks of denial are higher; coastal residents who deny climate change may find their land becoming seabed. Scientific evidence suggests that anthropogenic global warming acceptance in post-industrial nations has risen to as high as $81\%$ in the United States from near $0\%$ decades ago, contradicting the notion of an inflexible public.

The Primacy of Mechanistic Knowledge in Breaking Impasses

Not all knowledge is equal in its ability to influence belief. Ranney and Clark emphasize that mechanistic knowledge—understanding the "how" of a process—is paramount. In the context of global warming, mechanistic understanding identifies the causal relationships between sunlight, infrared light, Earth's surface, and greenhouse gases. This knowledge acts as a "tie-breaker" when information from authorities causes ambivalence. Without a mechanism, public discourse often devolves into debates over the authority or methodology of researchers. However, the normative scientific mechanism demands that a denier explain exactly why anthropogenic greenhouse additions are negligible or identify flaws in the chemical/physical consensus. Studies by Fernbach et al. support this, showing that soliciting mechanistic explanations reduces the "illusion of explanatory depth" and leads to more moderated, science-aligned attitudes.

Experiment 1: Measuring the Global Warming Wisdom Deficit

Experiment 1 consisted of a diagnostic survey of $270$ politically diverse individuals in San Diego and community college students. The sample was $39.3\%$ Democrat, reflecting national norms. The survey assessed policy preferences, belief in global warming, and mechanistic knowledge through a scoring rubric with high interrater reliability (\text{mean } j > 0.7). Results revealed a massive knowledge void: while $80\%$ accepted global warming, only $12\%$ had even a partial understanding (mentioning heat trapping), and only $3\%$ named the greenhouse effect. Crucially, only $1\%$ mentioned different types of light/energy, and $0\%$ mentioned the crux of the mechanism: the input/output asymmetry involving visible and infrared light. The median understanding score was $0$ and the mean was $0.65$ out of $3$. Misconceptions were common, with $74\%$ incorrectly blaming ozone depletion for global warming. Despite this ignorance, knowledge was significantly correlated with acceptance ($r = 0.22, p = 0.0002$) and willingness to sacrifice for the climate (\chi^2(4) > 32, p < 0.001).

Experiments 2, 3, and 4: Rapid Conceptual Change and Longevity

Experiments 2 through 4 tested the impact of a $400-word$ text explaining the global warming mechanism. Experiment 2 involved undergraduates from UC Berkeley ($N=85$) and UT Brownsville ($N=41$). Average knowledge scores doubled or tripled: understanding of light asymmetry rose from $20\%$ to $56\%$, and atmospheric energy-retention time understanding rose from $19\%$ to $48\%$. Global warming acceptance increased by $14\%$ of the "available room" toward extreme acceptance. Experiment 3 ($N=80$) replicated these findings online with a $4-day$ delay, showing that state global warming agreement rose from $6.20$ to $6.54$ on a $1-9$ scale and stayed at $6.44$ after the delay. Experiment 4 ($N=38$) used an Amazon Mechanical Turk (MTurk) sample and a $5.5-day$ delay. Scored knowledge jumped from $1.9$ to $4.8$ ($0-9$ scale), and acceptance increased from $6.3$ to $6.6$. Notably, there was no sign of polarization across the liberal-conservative spectrum ($r = -0.03, p = 0.85$).

Experiment 5: Classroom Curricula and 34-Day Retention

Experiment 5 expanded the intervention into a $45-minute$ high school chemistry curriculum ($N=63$). Two groups were compared: a mechanism-plus-statistics group and a mechanism-only group. The curriculum included simulations (PhET) and slide presentations. Pre-test knowledge was near zero. Post-test results showed marked improvement in both groups (p < 0.0001). However, the group that received six key statistics along with the mechanism showed significantly higher retention of mechanistic knowledge after $34\,days$ ($+0.8$ vs. $+0.3$ on a $0-4$ scale; t(48.7) = 2.61, p < 0.01). This suggests that specific statistical information reinforces and secures mechanistic understanding. Environmental Behavioral Intentions (EBI) also rose significantly ($5.7$ to $6.2$ on a $1-9$ scale).

Experiment 6: Numerically Driven Inferencing (NDI)

Experiment 6 evaluated the impact of statistical feedback in isolation using the Numerically Driven Inferencing (NDI) paradigm ($N=38$ MTurk workers). Subjects estimated seven representative quantities before receiving veridical feedback. Providing these statistics significantly increased global warming acceptance from $6.4$ to $6.8$ (t(37) = 2.74, p < 0.005). However, subjects reported a loss of self-rated knowledge confidence ($5.2$ falling to $4.2$) as they realized their prior estimates were wildly distal from reality. No polarization was observed ($r = -0.07, p = 0.67$), indicating that statistical evidence moves both liberals and conservatives toward the scientific consensus.

Experiment 7: Susceptibility to Misleading Statistics

Experiment 7 tested the effect of "cherry-picked" or misleading statistics used by agnotological groups. For example, subjects were told that Earth cooled by $0.2^{\circ}F$ between $1940$ and $1975$, a dip caused by temporary aerosol masking. In a group receiving eight such items ($N=59$), global warming acceptance significantly dropped by $11\%$ of the available room. Self-rated knowledge-confidence plummeted from $5.0$ to $2.9$, a $53\%$ decrease of the possible range. Funding preferences for global warming initiatives also dropped (p < 0.01) compared to other UN Development Programme (UNDP) goals. This demonstrates that scientific information is powerful regardless of its representative nature, and that the public is highly vulnerable to sophisticated distributions of out-of-context data.

Educational Outreach: HowGlobalWarmingWorks.org

To translate these findings into public service, Ranney et al. created HowGlobalWarmingWorks.org. The site features videos ranging from $52\,seconds$ to $4.7\,minutes$ ($83$ to $596\,words$). Key materials include the $35-word$ summary, translated Chinese (Mandarin) and German versions, and representative statistics and graphs. The goal is to provide accurate information directly to the public, bypassing journalists or teachers who may provide flawed or high-cognitive-load material (e.g., terms like "albedo" or "radiative forcing"). Preliminary data show that the $4.7-minute$ video can triple mechanistic knowledge and significantly increase acceptance.

Statistical Appendix: Key Quantitative Data and Summaries

The 35-Word Mechanism Summary: "Earth transforms sunlight’s visible light energy into infrared light energy, which leaves Earth slowly because it is absorbed by greenhouse gases. When people produce greenhouse gases, energy leaves Earth even more slowly—raising Earth’s temperature."

Experiment 6 "Representative" Statistics:

• Number of hottest years recorded between $1995$ and $2006$: $11\,of\,12$.

• Percent increase in atmospheric methane since $1750$: $151\%–150\%$.

• Percent decrease in world ocean ice cover since the $1960s$: $40\%$.

• Percent change in atmospheric $CO_2$ at Mauna Loa ($1959-2009$): $22.6\%–23\%–increase$.

• Percentage of climate researchers agreeing on human cause: $97.5\%$.

• Number of glaciers remaining in Glacier National Park (from $150$ in $1850$): $25\,glaciers$.

• Percent decrease in European Alps glacier volume since $1850$: $50\%$.

Aggregate Practical Gains: Interventions across Experiments 2–5 yielded a median knowledge gain of $41\%$ of the possible total ($28\%$ upon delayed post-testing). The median immediate acceptance gain was $14\%$ of what was possible, with a sustained gain of $9\%$ after delays ranging from $4$ to $34\,days$. Misleading statistics in Experiment 7 caused a $11\%$ drop in acceptance. Overall, the findings suggest that scientific information can durably transform attitudes and that the public's lack of knowledge is an opportunity for educational intervention rather than a state of permanent stasis.