More Like this

1. Surveys find "two kinds" of public knowledge about polar regions

   Hamilton, Lawrence C. ( November 2020 , Witness the Arctic)

   null (Ed.)

   How much does the US public know about polar regions? Researchers exploring this topic have occasionally mixed factual questions in among the more typical opinion queries on general-public surveys. A recent article in the Journal of Geoscience Education (Hamilton 2020) describes a key finding from these surveys: there are "two kinds" of polar knowledge. One kind is evoked by questions like this: Which of the following three statements do you think is more accurate? Over the past few years, the ice on the Arctic Ocean in late summer... - Covers less area than it did 30 years ago (correct) - Declined but then recovered to about the same area it had 30 years ago - Covers more area than it did 30 years ago The declining area of late-summer Arctic sea ice, tracked by satellites over the past 40 years, is a basic and widely reported scientific fact. On surveys, however, many people do not recognize this fact, but answer instead based on their opinion about global warming. Similar results occur if we ask whether, in recent decades, CO2 concentrations in the atmosphere have increased: again, many people give answers contrary to science, but reflecting instead their beliefs or political identity. Although these questions involve important and well-established facts, survey responses defy simple interpretation as indicators of knowledge. 

   more » « less
2. Motivating landowners to recruit neighbors for private land conservation

   https://doi.org/10.1111/cobi.13294  

   Niemiec, R_M ; Willer, R. ; Ardoin, N_M ; Brewer, F_K ( February 2019 , Conservation Biology)

   Encouraging motivated landowners to not only engage in conservation action on their own property but also to recruit others may enhance effectiveness of conservation on private lands. Landowners may only engage in such recruitment if they believe their neighbors care about the conservation issue, will positively respond to their conservation efforts, and are likely to take action for the conservation cause. We designed a series of microinterventions that can be added to community meetings to change these beliefs to encourage landowner engagement in recruitment of others. The microinterventions included neighbor discussion, public commitment making, collective goal setting, and increased observability of contributions to the conservation cause. In a field experiment, we tested whether adding microinterventions to traditional knowledge‐transfer outreach meetings changed those beliefs so as to encourage landowners in Hawaii to recruit their neighbors for private lands conservation. We delivered a traditional outreach meeting about managing the invasive little fire ant (Wasmannia auropunctata) to 5 communities and a traditional outreach approach with added microinterventions to 5 other communities. Analysis of pre‐ and post‐surveys of residents showed that compared with the traditional conservation outreach approach, the microinterventions altered a subset of beliefs that landowners had about others. These microinterventions motivated reputationally minded landowners to recruit and coordinate with other residents to control the invasive fire ant across property boundaries. Our results suggest integration of these microinterventions into existing outreach approaches will encourage some landowners to facilitate collective conservation action across property boundaries.

    

   more » « less
3. Unlearning Climate Misinformation in Large Language Models

   Fore, Michael ; Singh, Simranjit ; Lee, Chaehong ; Pandey, Amritanshu ; Anastasopoulos, Antonios ; Stamoulis, Dimitrios ( August 2024 , Association for Computational Linguistics)

   Misinformation regarding climate change is a key roadblock in addressing one of the most serious threats to humanity. This paper investigates factual accuracy in large language models (LLMs) regarding climate information. Using true/false labeled Q&A data for fine-tuning and evaluating LLMs on climate-related claims, we compare open-source models, assessing their ability to generate truthful responses to climate change questions. We investigate the detectability of models intentionally poisoned with false climate information, finding that such poisoning may not affect the accuracy of a model’s responses in other domains. Furthermore, we compare the effectiveness of unlearning algorithms, fine-tuning, and Retrieval-Augmented Generation (RAG) for factually grounding LLMs on climate change topics. Our evaluation reveals that unlearning algorithms can be effective for nuanced conceptual claims, despite previous findings suggesting their inefficacy in privacy contexts. These insights aim to guide the development of more factually reliable LLMs and highlight the need for additional work to secure LLMs against misinformation attacks. 

   more » « less
4. The causal impact of local weather anomalies on beliefs about the occurrence of climate change

   https://doi.org/10.1088/2515-7620/acffae  

   Carlson, Deven ; Ripberger, Joseph ; Jenkins-Smith, Hank ; Silva, Carol ; Carlson, Nina ; Bell, Elizabeth ; Gupta, Kuhika ( November 2023 , Environmental Research Communications)

   Research has demonstrated that members of the public recognize anomalous weather patterns, and that subjective perceptions of the weather are related to beliefs about the occurrence of climate change. Yet despite two decades of scholarship and dozens of studies, inconsistent and insufficient data have made it difficult to credibly identify the causal impact of objective experiences on perceptions, and the impact of perceptions on beliefs regarding climate change occurrence. Here, we overcome these limitations by collecting and analyzing data from a 5-y panel survey of 2,500 individuals in Oklahoma, a US state that is highly divided on questions about climate change. Our findings indicate that the relationship between local weather anomalies and climate change beliefs is heavily dependent on baseline beliefs about whether climate change was occurring. For people who did not believe in climate change in the initial survey in our series, perceptions of anomalously hot and dry seasons shifted their beliefs towards the occurrence of anthropogenic climate change, whereas their perceptions of anomalously cool and wet seasons shifted their beliefs away from anthropogenic climate change. This relationship was not present among people who believed that climate change was occurring at the beginning of the study; their perceptions of seasonal temperature and precipitation anomalies had no effect on their beliefs about climate change. These patterns have substantial implications for the evolution of public beliefs about climate change.

    

   more » « less
5. Six minutes to promote change: People, not facts, alter students' perceptions on climate change

   https://doi.org/10.1002/ece3.7553  

   Sauer, Kodiak A. ; Capps, Daniel K. ; Jackson, David F. ; Capps, Krista A. ( May 2021 , Ecology and Evolution)

   Anthropogenic climate change threatens the structure and function of ecosystems throughout the globe, but many people are still skeptical of its existence. Traditional “knowledge deficit model” thinking has suggested that providing the public with more facts about climate change will assuage skepticism. However, presenting evidence contrary to prior beliefs can have the opposite effect and result in a strengthening of previously held beliefs, a phenomenon known as biased assimilation or a backfire effect. Given this, strategies for effectively communicating about socioscientific issues that are politically controversial need to be thoroughly investigated. We randomly assigned 184 undergraduates from an environmental science class to one of three experimental conditions in which we exposed them to short videos that employed different messaging strategies: (a) an engaging science lecture, (b) consensus messaging, and (c) elite cues. We measured changes in student perceptions of climate change across five constructs (content knowledge, acceptance of scientific consensus, perceived risk, support for action, and climate identity) before and after viewing videos. Consensus messaging outperformed the other two conditions in increasing student acceptance of the scientific consensus, perceived risk of climate change, and climate identity, suggesting this may be an effective strategy for communicating the gravity of anthropogenic climate change. Elite cues outperformed the engaging science lecture condition in increasing student support for action on climate, with politically conservative students driving this relationship, suggesting that the messenger is more important than the message if changing opinions about the necessity of action on climate change is the desired outcome. Relative to the other conditions, the engaging science lecture did not support change in students' perceptions on climate, but appealing to student respect for authority produced positive results. Notably, we observed no decline in students' acceptance of climate science, indicating that none of the conditions induced a backfire effect.

    

   more » « less