Toronto Metropolitan University
Pretty_Brent.pdf (179.11 kB)

On Using Social Media to Predict Elections. A Meta-Analysis

Download (179.11 kB)
posted on 2021-05-24, 14:18 authored by Brent Pretty
As of late there's been great interest in social media’s ability to predict elections. Platforms such as Twitter and Facebook, owing to their cultural ubiquity, offer a plethora of data and an opportunity to track public perception at a granular level in real time. The ability to passively analyze public opinion is a massive step forward in the realm of political prediction, and has the potential to redefine the field of campaign strategy. In this piece of research I analyzed the current state of social media based electoral prediction. I examined the methods and techniques used to collect and analyze data, and compared their results against both each other and other methods of prediction such as telephone polling. In this I found a field that is still in its infancy. Much work remains to be done until a set of best practices surrounding social media based electoral prediction are accumulated.





  • Master of Digital Media


  • Digital Media

Granting Institution

Ryerson University

LAC Thesis Type

  • MRP

Thesis Advisor

Greg Elmer