Choice & Inference

Moshi Vardi asks, in the current issue of Communications of the ACM.
[Hat tip to Giuseppe Primiero]

Eric Gilbert, a 2009 Google Fellow, and his colleagues at Illinois Urbana-Champaign devised an empirical study to test whether blogs isolate readers from dissenting opinions by comparing the ratio of agreeing to disagreeing comments in several top blogs. They write that

As early as 1996, Nicholas Negroponte theorized about The Daily Me, a newspaper perfectly tailored to your individual tastes and preferences. Nothing appears in The Daily Me to challenge the beliefs you already hold. Cass Sunstein, a law professor at the University of Chicago, hypothesized that blogs may in fact be the modern Daily Me. Building on existing work in group psychology, Sustein warns that blogs acting as echo chambers could intensely polarize readers and snuff out dissent. Still, the question remains: are blogs echo chambers?

In their archly titled paper, Blogs are Echo Chambers: Blogs are Echo Chambers, Gilbert et. al. use a study of 75 face-to-face meetings which found that 18.7% of the time speakers actively agreed or disagreed with one another, and 64% of those opinionated moments were devoted to expressing agreement. They then categorized a blog as an echo chamber if more than 64% of the opinionated commentators agree with the blogger.

Their findings? In their hand-coded sample of over 1,000 comments from 33 of the world’s top blogs, agreement outnumbered disagreement by more than 3 to 1.