Authors: Drew Bush*, McGill University, Jin Xing, Newcastle University, Renee Sieber, McGill University
Topics: Environmental Perception, Communication, Quantitative Methods
Keywords: Social Media, Online Blogs, Word2Vec, Public Opinion, Climate Change
Session Type: Paper
Start / End Time: 3:05 PM / 4:45 PM
Room: Coolidge, Marriott, Mezzanine Level
Presentation File: No File Uploaded
The scientific community requires empirical evidence of public climate change opinions online. To date, public opinion research has focused on classifying segments of the public by their beliefs and determining how correcting and/or inoculating the public against misinformation can lead to informed public discourse. This manuscript examines how a data mining technique (Word2Vec) popular in the digital humanities can aid researchers studying public perceptions of climate change online. We sought to determine what role social media such as Twitter and online websites (such as those authored by climate scientists or professional journalists) played in civic discourse on climate change. We further sought to determine how climate related news or political events affected online discussions. By determining what words were most often associated with the term climate change on Twitter and four prominent climate science blogs, in June and September 2017, respectively, our method allowed us to characterize big datasets on each medium that involve large bodies of text and multiple posts from different users. Although it did not enable us to track how posts on either medium evolved in relation to each other over time, our findings confirm that climate news and political events can drive online conversation. These conversations were reflective of the culturally divisive political discourse on this topic, with the individual users’ opinions who we examined displaying varying degrees of understanding of climate change. We conclude with implications for journalists, scientists, and others working to correct public understandings of climate change.