Thursday, January 19, 2017

Stubbing my toe of perception on the rock of measured reality

From Pew Research Center, Trump, Clinton Voters Divided in Their Main Source for Election News: Fox News was the main source for 40% of Trump voters by Jeffrey Gottfried, Michael Barthel and Amy Mitchell.

A pretty good body of research with some interesting insights in terms of how people create their customized epistemic ecosystems. There are two observations not highlighted which seem important.

Click to enlarge

The headline leads with Fox News as a dominant source of news for those voting for Trump whereas Clinton voters had more diverse sources. The headline is "Fox News dominated as main campaign news source for Trump voters; no single sources pronounced among Clinton voters." True. But that seems to spin the information a particular direction (diversity good, monoculture bad) and it defines diversity as number of sources rather than orientation of sources.

An alternative way to look at this is that Republicans were more balanced in the ideological orientation of their news sources by balancing 40% from "conservative" news (from Fox) with 32% (omitting local radio) from "liberal" sources. In contrast, Democrats were much more in an echo-chamber with 74% of their news from "liberal" sources and only 3% from "conservative sources. In other words, Republicans favored their ideologically congruent sources by 25% (40/32 = 1.25) whereas Democrats favored their ideologically congruent sources by nearly a factor of 25 (74/3 = 24.7)

The alternative headline could be "Republicans dramatically more diverse in their news sources than Democrats."

Both statements, Pew's headline and the alternative I have suggested, are factually true. The only difference is how you measure diversity. Is it ideological diversity or is it source diversity? Both are interesting perspectives.

The second thing that leapt out at me was the paucity of three of my sources of information. The Washington Post does not show up at all. The New York Times and NPR only show up as relevant to Democratic voters and even then as relatively minor (5% and 7% respectively.)

I knew that all three, WaPo, NYT and NPR, were notably skewed in their political reporting and are consumed primarily by left of center college educated professionals. I knew they weren't directly significant to the public at large. I have shocked friends by pointing out that the audience of Rush Limbaugh (a conservative radio talk host with a two or three hour show) is nearly the same size as that of NPR's flagship news show, Morning Edition.

But knowing all that doesn't reduce the impact of seeing that for all voters, only 4% listed NPR as a main source of news and only 3% read the NYT as a main source of news.

Serves as an example that knowledge alone does not counter an anchoring bias. I consume all three sources (along with many others). I knew that their audiences were a fraction of the nation. And despite that knowledge, I imputed much more relevance to them than the measured reality supports.

No comments:

Post a Comment