Master's Thesis

Identifying Inequalities in Information Diffusion on Social Media: A Study of Twitter Cascades Across User, Content, and Contextual Groups

Tomás Manuel Estrócio e Silva — 2025

Key information

Authors:

Tomás Manuel Estrócio e Silva (Tomás Manuel Estrócio e Silva)

Supervisors:

Luís Humberto Viseu Melo (Luís Humberto Viseu Melo); Maria Joana Patrício Gonçalves de Sá

Published in

July 3, 2025

Abstract

How do user, content, and contextual characteristics influence cascade growth on social media? This study addresses that question by analysing over 1.9 million tweets and their retweets, considering user traits (gender, country, verification status, follower count), content veracity (true or false), and contextual factors (time of day). Cascade growth was modelled using an analytical function with two parameters: initial diffusion rate "a" and decay rate "g", both fitted to a Generalised Gamma Distribution (GGD). Group comparisons reveal clear asymmetries: verified users, men, and accounts with many followers tend to generate larger cascades. Female and unknown-gender users display statistically similar patterns. While true claims had heavier-tailed size distributions (indicating the presence of some very large events), veracity alone did not significantly increase expected retweets when controlling for other factors in regression models. Simulations using the fitted parameters replicated the overall cascade distribution but failed to reproduce group-level differences, likely due to the use of a fixed network structure. Regression models confirmed the strong effect of visibility on cascade size, but explained only a small portion of the variance in "a" and "g". This suggests that cascade dynamics (both early growth and later slowdown) are not fully explained by user, content, or contextual features alone, and may depend on other unobserved factors. These results show how exposure differences and social perception contribute to biases in information diffusion, especially by gender and visibility, offering ground for future research on algorithmic fairness and digital platform regulation.

Publication details

Authors in the community:

Supervisors of this institution:

Fields of Science and Technology (FOS)

physical-sciences - Physical sciences

Publication language (ISO code)

por - Portuguese

Rights type:

Embargo lifted

Date available:

May 2, 2026

Institution name

Instituto Superior Técnico