SoylentNews

Science of "Echo Chambers"

posted by cmn32480 on Thursday May 28 2015, @03:18AM   
from the hello-hello-hello dept.

aristarchus writes:

Oft times we see accusations of "group think" here on SoylentNews. Now there is some actual science on the formation and function of "echo chambers", as reported by SESYNC:

A new study from researchers at the University of Maryland (UMD) and the National Socio-Environmental Synthesis Center (SESYNC) demonstrates that the highly contentious debate on climate change is fueled in part by how information flows throughout policy networks.
...
"Our research shows how the echo chamber can block progress toward a political resolution on climate change. Individuals who get their information from the same sources with the same perspective may be under the impression that theirs is the dominant perspective, regardless of what the science says," said Dr. Dana R. Fisher, a professor of sociology at UMD and corresponding author who led the research.

I would guess, based on this study abstract (actual paper unfortunately behind paywall), that SoylentNews is in no danger of becoming an echo chamber, but we seem to have some refugees who are still stuck in particular bubbles.

• Rationale of Paper Rationale of Paper (Score: 0) by Anonymous Coward on Thursday May 28 2015, @02:32PM

  by Anonymous Coward on Thursday May 28 2015, @02:32PM (#189088)

  Interesting quote from paper:

  Although Fig. 2 shows the existence of some transitive triads in the network, descriptive measures alone cannot determine whether the number of transitive triads present could be explained by other network properties. For example, some such triads could be formed by chance, simply owing to the number of edges present in the network (note that the number of triads increases with the number of alters and ties in the ego network). Similarly, we must determine whether the empirical level of homophily observed-that an actor uses a source with the same opinion as him/herself-is also not due to random tie formation. Furthermore, to establish the presence of an echo chamber, our data must show, not only that both an echo and a chamber are operating, but that the transitive triads within which the information is transmitting are themselves sorted by homophily. In other words, we must show that echo chambers are statistically more likely to occur within like-minded, homophilous groups.

  First, I am more interested in talking about the line of reasoning than any of the conclusions. They do describe the rationale in some detail with the quote above, which is good.

  It is good that they recognize there may be other reasons for data consistent with an echo chamber than theirs. Then it goes downhill... the only other explanation considered is "chance". It is implausible there is zero relationship between beliefs and who you talk to, so this is a case where the chance explanation is a strawman.

  On the good side, they require further evidence to support the echo chamber idea. Specifically that the presence of these transitive triads (A told B and C, B also told C) is positively correlated with agreement on the various questions they asked. This is slightly better than ruling out the chance explanation, but really can you not think of one other possible explanation for this? The lack of even suggested possibilities brings the "science"-ness of this work into question.

• Re:Rationale of Paper Re:Rationale of Paper (Score: 3, Interesting) by aristarchus on Thursday May 28 2015, @05:50PM

  by aristarchus (2645) on Thursday May 28 2015, @05:50PM (#189216) Journal

  but really can you not think of one other possible explanation for this?

  Interesting critique. And possibly the elephant in the room. Besides chance and homophily (love that word, I may have to become a homophiliphobe!), the obvious explanation is intentional echo-chambering, based on private economic interest. In other words, shilling and astroturfing. It is so obvious, that the climate deniers turn it right around and accuse those who suggest action to counter global warming of having ulterior economic motivations. But maybe that is not what the study is attempting to understand. Maybe the real question is why such a tactic on the part of petrochemical corporations can actually work, how it is that normal people can be snookered into an echo chamber.

  That's enough argumentum ad hominem giganticam for one post. Elephant in the room, get it? Platinum bar!!

  Parent

  • Re:Rationale of Paper (Score: 0) by Anonymous Coward on Friday May 29 2015, @01:05AM

    by Anonymous Coward on Friday May 29 2015, @01:05AM (#189428)

    I'm thinking more input required: https://www.youtube.com/watch?v=Pj-qBUWOYfE [youtube.com]

    Parent

  • Re:Rationale of Paper Re:Rationale of Paper (Score: 2, Interesting) by Anonymous Coward on Friday May 29 2015, @03:54PM

    by Anonymous Coward on Friday May 29 2015, @03:54PM (#189712)

    Maybe the real question is why such a tactic on the part of petrochemical corporations can actually work, how it is that normal people can be snookered into an echo chamber.

    Well the answer to that is pretty obvious and even fairly well researched. A large proportion of the population, those normally self-identifying as conservative, are averse to change. Historically this has been a successful evolutionary strategy because useful change has been sufficiently rare that suggested deviation from tradition was more likely to be self serving or have negative impact. However the accelerating pace of scientific and technologic change has led to so many advantageous disruptions that the conservative, who normally are part of the successful group due to avoiding wasting energy on fads and dangerous trends, have found themselves left behind and disenfranchised instead. Anybody who provides an excuse for their continued resistance to change can find very fertile ground indeed for any meme complex which validates the conservative/change averse approach.

    Parent

    • Re:Rationale of Paper (Score: 0) by Anonymous Coward on Friday May 29 2015, @10:22PM

      by Anonymous Coward on Friday May 29 2015, @10:22PM (#189867)

      These are all interesting theories, but you have to show how it compares to the data in the paper.

      Parent