- R.Kelly Garrett. “Echo chambers online?: Politically motivated selective exposure among Internet news users.”
- Paul Resnick et al. “Bursting your (filter) bubble: strategies for promoting diverse exposure.”
Should we Care?
Yes this is an important topic. Facebook is now wrestling with how best to serve balanced news to its users after the fiasco that was 2016 cycle, together with the adverse effects of echo chambers: How do you satisfy opposing views in providing information that balances for both? And how to mitigate echo chamber’s nature of amplifying misinformation and hatespeech.
Glut vs Choice
Both papers are complementary: Garrett talks about how people consume news, and Resnick discusses how to nudge people towards consuming diverse news sources. They were written before the advent of #FakeNews, and so it does not account for the negative perception of news sources (together with the fast(er) news cycle. This glut of both news sources and news items: Social media posting links to outside news, serving on-demand video news, commentary, podcasts etc. Many more factors that fit the paper’s definition of news sources caliber would serve to give greater nuance and a new lens by which to review this work.
Skimming: then and now
I think we are past the era of news aggregation, aside from Google acting as a gate-keeper in providing a topical aggregation. There is a difference in how news is shared: Google/Twitter trending topics with teasers, the prevalence of paywalls on accessing online news paper, twitter brevity enforced by the character count etc., have led to increasing use of skimming. This runs counter to Kselly reasons this behavior to be due to prior knowledge of the news. It would be informative to see how their assumption scales to the new way news is consumed.
Variables Limitations
The dependent variable (Pop-up window) is inexact, though they account for extremes. Instead of considering active window time, scrolling behavior would’ve been a better judge of time use and provide granularity on reading behavior (i.e. skimming vs reading in depth).
On Independent variable (Perceptions on opinions): there was no delineation about political opinions formed by reading and those formed by personal belief systems that are unshaken and counters typical opinion formation. For example the author hints at gay marriage, at present, we have the issue of abortion that transcends politics and into what individuals believe in their core. It was not clear from the paper whether they considered this a lumped category. But a gradient would be useful first in separating opinions from beliefs and second, in ascertaining how best to present nudging opportunities depending on scale.
The users used in Kelly’s work were partisan. While this allowed for a good study of contrast — the authors also noting in the limitations about how this skew unnaturally compared to the general population, it also provides an inspiration on where to start/proceed with measurements: They claim that there’s an increased awareness of politics as we get older, how much this translates to the general population would also be more useful knowledge towards using for nudging/intervention.
Other consideration/measurements
- Whether in the process of posting/rebutting opinions the echo chamber formation also extends to the kind of friends that interact/surround an individual.
- What is the ethics of nudging? Knowledge towards diverse source can also be used to lead other people astray.
