The purpose of this research is to detect and quantify echo chambers on social media and to explain the mechanisms that sustain them. I will answer this main question by measuring how topic homogeneity, sentiment uniformity, and network connectivity co-occur inside detected communities. As I conduct this research, I also want to find out which combinations of content and network features correspond to the strongest echo chambers and if echo-chamber characteristics differ between Reddit and X for the same event.

I will be building visual graphs to analyze communities within social media platforms as well as calculating an echo chamber index. In the graph, communities will be groups of comments that discuss the same topic. The index will be built based on three categories. The first is network isolation. This is to see how closely related comments are to other comments in its community as opposed to other comments in external communities. The next is sentiment similarity. If all comments in a community feel the same way about the respective topic, then I will treat that as more of an echo chamber. Lastly, I will look at topic similarity. The more closely related topics are from each comment, the more interconnected an community would be, making it more of an echo chamber.