Misinformation and/or Disinformation on Climate Change By Internet Influencers

Journal for High Schoolers, Journal for High Schoolers 2022

Mentor: Amy Dunphy

By: Alejandro Macias, Krish Dev, Nhi Huynh, Corina Chen, Michael Sun

Abstract:

Disinformation and misinformation on climate change have been present early on as the discovery of human cast-offs led to the rise of carbon dioxide emissions and thus global warming. As early as 1980, oil industry companies, such as the American Petroleum Institute, had propagated and spread disinformation to the general public about climate change in order to lean the general public’s attention towards the fossil fuel industry.

Adding to present research, our work seeks to analyze the connection between misinformation and disinformation about climate change and the influence that has spread it and/or started it.

Background:

Over a seven-year period from 2013 to 2020, Twitter was recorded to have received an increase of an average of 50% in the use of the hashtag #climate change. According to Twitter, the discourse around climate change has risen significantly throughout the years. Twitter debates about the climate problem are increasingly being influenced by artificial bots. Automated bots are responsible for a quarter of all claims related to the climate crisis. Some topics have a higher percentage, like 38% of tweets about “fake science” and 28% of all tweets about ExxonMobil, which is an oil company with a history of climate denial. In an effort to dismantle the incorrect information about climate change, Twitter has implemented a new prohibition on advertisements that are misleading and contrary to the scientific consensus on climate change. Twitter’s announcement is also part of a broader social media story aimed at stopping misinformation about climate change.

In term of the consequences that misinformation about climate change has, inaccurate information regarding climate change has misled the public, prevented politicians from taking action, and prevented or slowed support for mitigation programs. It already has been a big contributor to slow or no progress in fighting climate change. First, a persistent, decades-long campaign of disinformation is a significant factor in the public’s false conceptions about climate change. It reduces acceptance of climate change and trust in people’s knowledge. Secondly, conservatives are also disproportionately affected by climate disinformation, which has exacerbated divisiveness in recent years.

The aim of this research is to discuss online misinformation, and how it pertains to climate change, and address the following key questions (1) How has social media, specifically Twitter, impacted climate change, changing the narrative of climate activism and spreading the correct information on the topic? (2) Does Twitter significantly play a role in improving climate literacy, especially in this modern and technological world? (3) How does the media underreporting of climate change affect climate action?

Methods:

We first created Twitter API developer accounts, which give us access to data, actions, and other activities that a user can do on Twitter. In order to open a Twitter API developer account, it acquires an API key, an API token, and bearer token using tokens. With a Twitter API developer account, we can download tweets onto our computer for analysis. After that, we started to look into tweets, public figures, and climate change activists who actively engage in the climate change and climate awareness conversations on Twitter and store them in Docs. We then use Tweetdeck to look for tweets and hashtags that are frequently repeated or generated by automated bots. Like Twitter, Tweetdeck has a search function with various columns, so many searches could be conducted at the same time. After that, we use Tweepy in Python to collect tweets related to climate change and analyze how many times different terms were repeated in 100 tweets related to a specific hashtag. Tweepy is a Python library which helps us access the Twitter API, which in turn helps us retrieve and engage with the data within the Twitter space.

With the use of the library, we used functions accordingly to scrape the data from the Twitter API. As the final step, we use python to create different visualization graphs to illustrate the data that we collected.

Image 1
Image 2

With the first image, we were able to filter through millions of tweets and collect 100 that use the keyword, “#climate”. We then scraped the tweets into the terminal allowing us to more efficiently use them, as seen in the second image

Image 3
Image 4

We are then allowed to analyze the tweets with the key words, “no climate change,” and “lie” to better narrow down our findings.

Conclusions:

This paper has presented the way social medias impact the way we think about the information on climate change, specifically on twitter. Through the means of the Twitter API and Tweepy, we were able to find the effects of climate activism and how ‘bots’ interact with them. When it comes to the effects of spreading false information on climate change, it has misled the general public, stopped governments from acting, and slowed down or stopped funding for mitigation efforts. In addition to this it decreases belief in people’s knowledge and tolerance of climate change. We eventually concluded that social medias such as twitter play a vital role in climate literacy in the technological world we are in and its underreporting is something everyone should take a look at.

As a team we experimented with new resources such as tweepy, pytorch, and much more to scrape tweets into our data, create visualization graphs, and analyze our tweets to narrow down our findings. With certain hashtags such as ‘#climate_emergency’ and ‘#climate_crisis’ we were able to create conclusions on the misinformations and where it all stems from. Looking at the tweets that we pulled, the comments had terms in denial including “joke,” “scam,” and etc. With all of this being said, we are able to create a more informed judgement on this issue and advise others as well to do the same.

Current Results (Krish)

Future Directions:

Given by the results, we can infer that this specific method of sentimental analysis can help us achieve similar results for other applications in other industries, such as politics (measuring how many times a candidate mentions a specific term) and advertisement (calculating the effectiveness of the propaganda analytics.) This can provide us with better insights to formulate better strategies to improve the world in various ways by optimizing specific operations.

With that being said, in order to achieve better results that are more accurate without error and noise, we have to improve our calculation and retrieval method on a more rigorous manner. This may include, but not limited to, collecting more datapoints, increasing the filters during our word collection process to uplift the bar of our criteria for key words that can possibly eliminate further noise within the data, and incorporate the other relevant information that can provide more context to the certain words that we may not have included in the original data retrieval.

The new improvements can also hope to bring new visualizations to our data that may be more helpful to provide more insights. By creating new models, such as linear regression or data classification techniques, we may be able to discover a new relationship with the help of the visuals that was not as evident theoretically.

References:

Brulle, R. J. (2018, July 19). The climate lobby: A sectoral analysis of lobbying spending on climate change in the USA, 2000 to 2016 – Climatic change. SpringerLink. Retrieved August 5, 2022, from https://link.springer.com/article/10.1007/s10584-018-2241-z

Explore brown university. Climate Science and Twitter Bots | Data Science | Brown University. (n.d.). Retrieved August 5, 2022, from https://www.brown.edu/initiatives/data-science/news/2020/03/climate-science-and-twitter-bots

Guardian News and Media. (2022, April 23). Twitter uses Earth Day to announce ban on climate denialism ads. The Guardian. Retrieved August 5, 2022, from https://www.theguardian.com/technology/2022/apr/23/twitter-bans-ads-deny-climate-crisis

Raj, A. (2021, September 27). How twitter hashtags can help understand climate change. Tech Wire Asia. Retrieved August 5, 2022, from https://techwireasia.com/2021/09/how-twitter-hashtags-can-help-understand-climate-change/

Leave a Reply