Torrential Twitter: Measuring the Severity of Harassment when Canadian Female Politicians Tweet about Climate Change
The online harassment of female politicians who focus on climate change and environmental policy has become a major problem in Canada and other democratic nations. Despite growing awareness of the problem, there is little agreement among scholars on how to measure these nuanced forms of harassment. This study develops an original seven-point scale to measure the severity of harassment three Canadian female politicians receive when Tweeting about climate change, and a six-point scale to categorize the types of accounts behind the replies. My results reveal that 86% of replies contained some form of harassment, most often name-calling or questioning the authority of the female politicians and come from users with spam or anonymous accounts. Further results from my Bayesian hierarchical model suggest that despite differences in status and political affiliation across the three politicians, they are almost equally impacted by harassment when Tweeting about climate change. These findings contribute to understanding the intersection between climate change denialism and the gendered nature of online harassment.This paper contains language and themes that some readers may find offensive.
The repo is structured as the following:
-
Inputscontains the PDF of the Codebook to code the Tweets. -
Outputscontains a sample dataset and the files used to generate the paper, including the Quarto document, bibliography file, model, and the PDF of the paper. The presentation slides and files used to generate them are also in this folder. -
Scriptscontains the R scripts used to simulate, download, clean, validate, and model the data.
The raw and cleaned datasets cannot be uploaded to GitHub. A sample dataset of the coded Tweets without any identifying information can be found in the Outputs\Data folder. Please contact Inessa (Inessa.DeAngelis@mail.utoronto.ca) to access the full datasets.