computational-sociology

What problem does it solve?

It helps you turn messy social-media and survey data into measurable evidence about social networks, sentiment, polarization, and causal preferences.

Core Features & Use Cases

• Social-media data collection: Pull Twitter/X posts via Academic API and Reddit posts via PRAW for downstream analysis.
• Text preprocessing + sentiment scoring: Clean social text and compute VADER sentiment scores to quantify affect over time or topics.
• Bot detection and filtering: Apply heuristic bot-score rules to reduce low-quality or non-human accounts before analysis.
• Network analysis for homophily and echo chambers: Compute assortativity-based homophily and a random-walk segregation score for echo chambers.
• Conjoint experiment inference: Estimate AMCEs (Average Marginal Component Effects) and marginal means from choice-based conjoint datasets.

Quick Start

Use the computational-sociology skill to collect tweets and Reddit posts for your target period, compute sentiment, filter suspected bots, and then measure homophily and echo-chamber segregation from the resulting network.