Causal Inference in Communication Studies
date
Jun 27, 2024
slug
causal-inference-in-communication-studies
status
Published
summary
Causal inference has long been a prevalent methodology in political science and economics, but its adoption within the field of communications has been slower, albeit increasing. This post aims to gather pioneering studies that utilize causal inference within the field of communications.
tags
Academic
Communication
Data Analysis
Methodology
Paper Reading
type
Post
Causal inference has long been a prevalent methodology in political science and economics, but its adoption within the field of communications has been slower, albeit increasing. This post aims to gather pioneering studies that utilize causal inference within the field of communications.
Recent Studies Using Causal Inference
Civilizing social media: the effect of geolocation on the incivility of news comments
New Media & Society
2023
Yufan Guo
Yuhan Li
Tian Yang
RDD
ITS
Privacy cynicism and diminishing utility of state surveillance: A natural experiment of mandatory location disclosure on China’s Weibo
Big Data & Society
2024
Yuner Zhu
ANNOVA
Post-January 6th deplatforming reduced the reach of misinformation on Twitter
Nature
2024
Stefan McCabe
Diogo Ferrari
David Lazer
Kevin Esterling
Jon Green
DiD
Reporting after removal: the effects of journalist expulsion on foreign news coverage
Journal of Communication
2024
Matt DeButts
Jeniffer Pan
GSC
Common Strategies in Causal Inference
- Difference-in-Difference (DiD)
- Overview: compare the difference between the differences between before and after the treatment within two groups
- Assumptions
- Derivatives
- Regression Discontinuity Design (RD, RDD)
- Overview
- Assumptions
- Derivatives
- Synthetic Control Method (SC, SCM)
- Overview: create a control group after treatment similar to all control groups before the treatment
- Assumptions
- Linear correlation
- Derivatives
- Instrumental Variable (IV)
- Overview: use an exogenous variable which is correlated with X but independent on Y to estimate in two regressions
- Assumptions
- Derivatives