We are analysing the online interactions between audiences and media in the context of journalist practices during the mass protests in Chile in the second half of 2019. The relevance of these mass protests is associated with the opportunity window to draft a New Constitution in Chile after the social unrest.
We focus on the activity on Twitter of 22 legacy and non-legacy Chilean media, therefore, using the Twitter API, we have already downloaded more than 20 million tweets published between July 2019 and January 2020. Our empirical strategy relies on difference-in-differences regressions to estimate the effect of the type of media and coverage on the levels of interactions such as replication, replies and citation during the onset of mass protests in October 2019. For our descriptive analysis and estimations, we are applying a novel algorithm to classify the toxicity levels in the messages, particularly the replies to the content published by the media in our sample.
|1. Data Gathering||2. Data Cleaning|
|3. Sentiment Analysis||4. Toxicity Algorithm|
Figure by Orchard, González-Bustamante and Cisternas (2023).
Last updated: July 6, 2023.