Sunday reading -- & listening: the "backfire effect"

There's been some interesting discussion in the "comments" field (here, e.g.) on whether factual corrections "backfire," thereby entrenching mistaken beliefs that are ideologically congenial.

So here are a couple more logs for the fire on that.

1st, an interview of Brendan Nyhan, who speculates on whether his own studies suggesting a backfire effect might be wrong ("Walking back the backfire effect"). His willingness to question his own work displays admirable scholarly chracter.

2d, a new empirical study:

I haven't yet read it closely -- hope to later today -- but am curious to see what others think of it.

Comments

[Jonathan]

The backfire effect is undead!

It would be nice if they tested the salience effect directly. For instance, if they had another treatment group that read an item mentioning immigration without anything that might bring to mind MLP's alt-facts. Perhaps a positive item about immigrants. If reading about an anecdotal positive immigrant story raises support for MLP, then that really does suggest a salience effect. Like what happens to me (deathly allergic to cats) when I watch an internet cat video (you think they're cute, but I know they are EVIL!).

Which brings me to: I wonder if immigration, because it perhaps triggers behavioral immune system responses (and a corresponding salience-driven backfire effect - like how just thinking about cats has made me crank up my air filter to 11, even though no cats are around), is different from issues discussed frequently here (climate change, vaccines, guns, etc.) that probably don't interact with issue-specific evolutionary cognition.

However, note that Wood & Porter did use two immigration-related statements (one about Mexican immigrants and crime, the other about whites becoming a minority), and neither demonstrated a direct belief backfire effect. Maybe they would also have demonstrated an electoral preference backfire effect?

Also note that MLP's alt-facts were not completely contra-facts. Adult males were disproportionately represented in factual immigration figures, and they have a slightly higher real unemployment rate, and there was a famous French resistance in WWII. Just not close to MLP's claims (for the first two - she didn't explicitly give any numbers for the last). Maybe hyperbole is perceived as a positive attribute? But an effect like that might have shown up as well in Wood & Porter's white-minority statement, but didn't (although, again, perhaps because they didn't measure electoral preference).

Lastly, because this paper shows a disconnect between belief revision and electoral preference revision, I wonder if the cognitive disconnect is even more thorough - as in the voter choice blindness studies like DOI 10.1371/journal.pone.0060554 vs. DOI 10.1371/journal.pone.0171108. If one discounts voters that are choice blind to their electoral preference, does that eliminate many that have a belief/preference disconnect? Or, more frighteningly, does it eliminate many that maintain belief/preference consistency?

[Andy West]

Misinformation and Its Correction: Continued Influence and Successful Debiasing, 2012.
Stephan Lewandowsky, Ullrich K. H. Ecker, Colleen M. Seifert, Norbert Schwarz, and John Cook.

...claims a backfire effect when attempting to correct misinformation, referencing various other subsidiary sources.

[Jonathan]

"Beyond Counting Climate Consensus":
http://www.tandfonline.com/doi/full/10.1080/17524032.2017.1333965

[Joshua]

Jonathan -

Interesting article, but.... hmmm...

Drawing on examples from the extensive science and technology studies (STS) literature, we show that building the basis for policy action cannot be done simply with appeals to fact.

It think that such casual treatment of accuracy undermines their insight into effective communication.

[Joshua]

Oy. This is also rather dubious.

Rather, these arguments demonstrate the pitfalls of attempting to quantify consensus in the scientific literature in the manner of C13 in order to produce “proof” for persuading the public. Rather than securing certainty that was absent before, this exercise has invited intense scrutiny to the judgments underpinning their claim, and generated further doubt. This was a predictable outcome on the basis of STS studies which show that doing more research on politically controversial, high-stakes policy matters typically increases uncertainty (Collingridge & Reeve, 1986 Collingridge, D., & Reeve, C. (1986). Science speaks to power: The role of experts in policy making. New York, NY: St. Martin’s Press.

I haven't chased down their references, but I suspect that they have not sufficiently cited solid enough evidence so as to distinguish between "provokes criticism " and "increases uncertainty."

IMO, I have seen this same problem crop up many times. They are a assigning a causal role for consensus messaging to create backlash from those who are ideologically aligned against both the messaging and the messengers. What is truly causal there? How is the difference between consensus messaging acting as a mediator or a moderator established?

'Fraid to say, this looks like an article written to confirm a bias.

[Jonathan]

Joshua,

I'm not a fan of that article myself. I think the article desperately needs a post post-modernist editor.

Here's the science communication about this article:
https://phys.org/news/2017-07-campaigning-climate-science-consensus-backfire.html

[Ecoute Sauvage]

This is the opening sentence of the linked article:

"How persuasive are “alternative facts,” i.e., false statements by populist politicians, in convincing voters? "

Proof of those alternative facts would never convince these 4 article authors? By their definition, such proof is impossible, since only their definition is true. From the gun-control lobby we have an example for "secure" guns, when any hacker knows the alternative facts are demonstrably the true facts:

https://www.wired.com/story/smart-gun-fire-magnets/
>>>>>>>>>>>>> Gun rights advocates have long opposed smart guns, seeing them as a Trojan Horse to regulate traditional firearms. And Plore says his interest in smart-gun hacking originated with reading a vehemently anti-smart-gun thread on the gun forum Calguns.net in 2015. "Could you imagine what the guys at Defcon could do with this [piece of shit]?" one poster wrote wrote in response to a negative NRA review of the Armatix IP1.<<<<<<<<<<

Re-drafting the definition of "alternative facts" as Nyhan (to his credit, as Dan notes) is doing, is the obvious next step. Doubling down on demonstrable falsehoods or at best exaggerations simply discredits all other statements from the same source - how is this not obvious?

[Ecoute Sauvage]

Another "alternative fact" proven true is wonderfully illustrated in the new cover of Der Spiegel.
Title: Das Kartell

>>>>>>>>>>>>>>> https://magazin.spiegel.de/SP/2017/30/ <<<<<<<<<<<<<<<<<<<

What is this cartel? Why, only the engineers of the German car companies, who knew full well that efforts to comply with the Kyoto protocol's promotion of diesel engines rather than gasoline - in order to decrease the dreaded CO2 - can only succeed by a heroic fudging of actual emissions. So those poor guys are doing their level best to comply with their government's stated policy, and suddenly they're branded as vile conspirators? More alternative facts!