Homophily and influence
See also Causality, Causal inference, Causal inference with observational data, and Contagion.
It has been well-known that “similarity breeds connection”[^1], and the phenomenon is called ‘homophily’. One of the biggest issues is “can we distinguish influence from homophily with observational data?” James Fowler and Nicholas Christakis published a series of paper on the spreading of obesity, smoking, etc through a social network (most famously about the obesity; Christakis2007spread).
There has been a long debate on these papers. Strong rebukes such as Lyons2011spread have been published. In particular, Shalizi2011homophily demonstrated that homophily and contagion are generally confounded and are extremely difficult to tease out due to the latent homophily.
Articles
- http://bactra.org/notebooks/homophily-vs-influence.html
- Is obesity contagious? A Review of the Debate over the “Network Effects” of Obesity by Conrad Lee
- Social contagion? Maybe not… by Michał Bojanowski
References
- Quantifying causal influences
- Feedback effects between similarity and social influence in online communities
- Everyone’s an influencer: quantifying influence on twitter
- Differences in the mechanics of information diffusion across topics: idioms, political hashtags, and complex contagion on twitter
- TwitterRank: finding topic-sensitive influential twitterers
- Measuring user influence in Twitter: The million follower fallacy