This is the repo to replicate the results in:
Navelski, Joseph and Badruddoza, Syed and J. McCluskey, Jill and G. Pascual, Francis, Applied Network Science, 2026 https://link.springer.com/article/10.1007/s41109-025-00759-y#citeas
Supplementary Materials: (Springer PDF)
Users are encouraged to reach out to the corresponding author if they have questions about using the code.
We develop a model to infer the ideological positions and social influence of agents using social media data. Unlike existing approaches, our framework requires only a subset of network connections, rather than the full network structure and detailed socio-demographic characteristics of agents.
We apply the model to Twitter data to analyze interactions between experts in genome editing in livestock (GEL) and their followers. The model produces ideological position scores for both experts and followers, along with measures of their relative influence.
Our estimates indicate that followers opposed to genome editing exert greater influence than those in favor; for example, anti-GEL followers account for approximately 69% of the total social influence in a typical conversation. A post hoc analysis further suggests that the consensus emerging from interaction and learning dynamics is opposed to GEL.
More broadly, the model provides a flexible tool for inferring positions, perceptions, and influence in social media networks across topics, with implications for investment, marketing, and policymaking in the livestock sector.