Allaxys Communications --- Transponder V --- Allaxys Forum 1
Aktuell im WWW => *** PRESSEMELDUNGEN *** => Topic started by: Krik on January 22, 2020, 05:24:20 AM
-
[*quote*]
Consumer Health Digest #20-03
January 19, 2020
Consumer Health Digest is a free weekly e-mail newsletter edited by William M. London, Ed.D., M.P.H
http://www.calstatela.edu/faculty/william-m-london
., with help from Stephen Barrett, M.D
http://www.quackwatch.org/10Bio/bio.html
It summarizes scientific reports; legislative developments; enforcement actions; news reports; Web site evaluations; recommended and nonrecommended books; and other information relevant to consumer protection and consumer decision-making. Its primary focus is on health, but occasionally it includes non-health scams and practical tips.
###
Couple suing promoters of chiropractic neuropathy treatments
In the fourth of a series of investigative reports about a Southern California-based chiropractic business that aggressively markets non-validated neuropathy treatments, the NBCLA I-Team has told the story of patients Donna and Harvey Stone who say:
* They were duped into financing treatments by a sales presentation at a free dinner they attended after receiving an invitation.
* They did not understand that the doctors offering treatment were chiropractors when they first arrived at the business.
* They had a total bill of $18,000, none of which is covered by insurance.
* Harvey received a severe burn on his leg from equipment used to treat him and required weeks of treatment from a wound care facility.
The Stones are suing Optimal Health/Straw Chiropractic and its owner, chiropractor Philip Straw. Attorneys for the Stones want other patients to know that they likely have legal recourse too. Optimal Health/Straw Chiropractic shut down in January 2019, but shortly afterward opened as Superior Health Centers, which appears to use identical marketing materials.
[Johnson C. Roher C. Couple accuses SoCal chiropractic business of swindling them out of $18K
https://www.nbclosangeles.com/investigations/couple-accuses-socal-chiropractic-business-of-swindling-them-out-of-18k/2291865/
. NBC Los Angeles. Jan 16, 2019]
The lawsuit is scheduled for trial in February. The complaint
http://www.casewatch.net/civil/straw/second_amended_complaint.pdf?_ga=2.235929353.969045518.1579461064-1489850722.1556638581
alleges medical negligence in the treatment of Harvey, elder abuse/financial abuse, negligent infliction of emotional distress, fraud/intentional misrepresentation, and concealment.
###
Concerns that factual arguments will backfire allayed
Full Fact, the UK's independent factchecking charity, has published a brief report which concludes that "factchecking does help to inform citizens and backfire effects are rare rather than the norm. We still need more evidence to understand how factchecking content can be most effective."
[Sippit A. The backfire effect. Does it exist? And does it matter for factcheckers?
https://fullfact.org/media/uploads/backfire_report_fullfact.pdf
Full Fact, Aug 2019]
The report notes:
* A "backfire effect" refers to the effect that, when a factual claim reinforces someone's ideological beliefs, telling them that the claim is wrong ("debunking" it) can actually make them believe the claim more strongly rather than less thereby meaning that factchecking is ineffective or even counterproductive.
* Seven major experimental studies have examined supposed backfire effects.
* While backfire may occur in some cases, the evidence now suggests it is rare rather than the norm, and that debunking can make people's beliefs in specific claims more accurate.
* Two studies that tested the effects of addressing specific fears about vaccines found that debunking were effective at reducing beliefs in the claim they were debunking, but still decreased the intention to vaccinate among those with high levels of concern about vaccine side effects.
###
Alleged pyramid scheme temporarily shut down
A federal court has granted the Federal Trade Commission's (FTC's) request to temporarily shut down
https://www.ftc.gov/system/files/documents/cases/netforce_seminars_tro.pdf
an alleged pyramid scheme known as "Success By Health," and to freeze the assets of the company and its executives.
[FTC acts to shut down 'Success by Health' instant coffee pyramid scheme
https://www.ftc.gov/news-events/press-releases/2020/01/ftc-acts-shut-down-success-health-instant-coffee-pyramid-scheme
. FTC press release. Jan 16, 2019]
In its complaint
https://www.ftc.gov/system/files/documents/cases/netforce_seminars_stamped_complaint.pdf
, the FTC alleged that Success By Health and its executives James "Jay" Dwight Noland, Jr., Lina Noland, Scott A. Harris, and Thomas G. Sacca:
* operated a pyramid scheme that used false promises of wealth and income to entice thousands of consumers to join
* featured an instant coffee called "MycoCafe" that included a mushroom that the defendants claimed has health benefits, but selling the product to coffee drinkers took a back seat to recruiting more affiliates
* have taken more than $7 million from consumers and pocketed over $1.3 million for themselves
* promised consumers "financial freedom," even though fewer than 2% of participating consumers received more money from the defendants than they paid to them; and that those lucky few averaged less than $250 per month
* told affiliates that "the masses" could earn more than $1 million each month in sales commissions, but marketing materials failed to disclose that to achieve that level of commissions, an affiliate would have to recruit more than 100,000 affiliates working underneath them, the vast majority of whom would be losing money at any given time
The FTC also alleges:
* The company also sold its products directly to the public for the same "wholesale" price paid by affiliates, severely limiting affiliates' ability to follow the defendants' instructions to apply a 50% "markup" before selling to the public. Thus when affiliates tried to sell the product to other consumers, they found themselves in competition with the company itself.
* Jay Noland is violating a 2002 court order stemming from a previous FTC case related to another failed pyramid scheme
https://www.ftc.gov/news-events/press-releases/2002/07/pyramid-promoter-settles-ftc-charges
###
Multilevel marketing blasted
Business writer Mona Bushnell has noted that many (and some would say nearly all) multilevel marketing companies operate as pyramid schemes in which salesperson make money primarily by recruiting more and more downline sales associates, not by direct sales of products.
[Bushnell M. MLMs are preying on the dream of entrepreneurship
https://www.business.com/articles/mlms-target-women-and-immigrants/
. Business.com. Aug 22, 2019]
She asserts that:
* Pyramid schemes prey on women, minorities and immigrants.
* Pyramid schemes use the language of success.
* Direct sales companies purposefully obfuscate sales data.
* The viciousness of multilevel marketing (MLM) companies cannot be overstated, no matter how family-friendly, female-friendly, and diverse they might seem.
=================================
Stephen Barrett, M.D.
Consumer Advocate
287 Fearrington Post
Pittsboro, NC 27312
Telephone: (919) 533-6009
http://www.quackwatch.org (health fraud and quackery)
[*/quote*]
-
This made me raise my eyebrows. They make studies? About what!?
Fact is, and this we see day for day: Their "debunking" is no real debunking, but only plain hogwash. That hogwash can not convince anyone. Except for those people, who parrot what they are told. Parrots are parrots. Don't care about the fraction of parrots. Do concentrate on the majority.
The majority OF COURSE can be told facts. But they must be told in an intelligent way. No hogwash, but plain facts. FACTS!
That is what we do. We do not give a shit on "backfire". We tell plain facts. No one does it better than us. That is why the mystics and other criminals are so afraid of us. 8)
[*quote*]
Concerns that factual arguments will backfire allayed
Full Fact, the UK's independent factchecking charity, has published a brief report which concludes that "factchecking does help to inform citizens and backfire effects are rare rather than the norm. We still need more evidence to understand how factchecking content can be most effective."
[Sippit A. The backfire effect. Does it exist? And does it matter for factcheckers?
https://fullfact.org/media/uploads/backfire_report_fullfact.pdf
Full Fact, Aug 2019]
The report notes:
* A "backfire effect" refers to the effect that, when a factual claim reinforces someone's ideological beliefs, telling them that the claim is wrong ("debunking" it) can actually make them believe the claim more strongly rather than less thereby meaning that factchecking is ineffective or even counterproductive.
* Seven major experimental studies have examined supposed backfire effects.
* While backfire may occur in some cases, the evidence now suggests it is rare rather than the norm, and that debunking can make people's beliefs in specific claims more accurate.
* Two studies that tested the effects of addressing specific fears about vaccines found that debunking were effective at reducing beliefs in the claim they were debunking, but still decreased the intention to vaccinate among those with high levels of concern about vaccine side effects.
[*/quote*]
https://fullfact.org/media/uploads/backfire_report_fullfact.pdf
[*quote*]
The backfire effect
Does it exist? And does it matter for factcheckers?
Amy SippittFull Fact is the UK’s independent
factchecking charity.
About this paper
Full Fact’s Research Manager, Amy Sippitt, studies what
existing academic evidence tells factcheckers about how our
work can be most effective. This briefing looks at the common
question posed to factcheckers of whether there is a backfire
effect.
We would like to extend our warmest thanks to Brendan
Nyhan, Ethan Porter and Briony Swire-Thompson for their
comments.
Published by Full Fact, March 2019
Full Fact
2 Carlton Gardens
London
SW1Y 5AA
team@fullfact.org
+44 (0)20 3397 5140
@FullFact
fullfact.org
Registered Charity number 1158683
Published under the Creative Commons Attribution-ShareAlike
4.0 International License
2
Full Fact The backfire effect: does it exist? And does it matter for factcheckers?Contents
Quick summary 4
Introduction 5
How we’ve selected the studies, and some questions you might consider 8
So, where has backfire been observed? 9
11
Other studies that have found some form of backfire
Recent attempts to replicate have found no backfire 13
Other studies 20
Studies in the UK 20
Studies looking at broader issues 20
What does this all mean for factcheckers?
3
Full Fact The backfire effect: does it exist? And does it matter for factcheckers?
21Quick summary
• A “backfire effect” refers to the effect that, when a factual
claim reinforces someone’s ideological beliefs, telling them
that the claim is wrong (“debunking” it) can actually make
them believe the claim more strongly rather than less.
• This supposed effect is often interpreted as meaning that
factchecking is ineffective, or even counterproductive.
• This briefing looks at seven major experimental studies that
have examined supposed backfire effects, mostly in the
United States.
• It finds that while backfire may occur in some cases, the
the evidence now suggests it is rare rather than the norm,
and that generally debunking can make people’s beliefs in
specific claims more accurate.
• Two studies, from 2010 and 2012, found some evidence of
a backfire effect in certain circumstances.
• None of the five more recent studies looked at (from 2015,
2017, 2018 and 2019) have found any evidence of the
effect.
• The cases where backfire effects were found tended to be
particularly contentious topics, or where the factual claim
being asked about was ambiguous.
• Our worldview may still affect the extent to which a debunk
might be effective.
• The impact of debunks on our behaviour is more
complicated. This briefing does not look at this in detail; it
focuses on how debunks affect belief in the accuracy of the
claim in question.
• This review suggests factchecking does help to inform
citizens and backfire effects are rare rather than the
norm. We still need more evidence to understand how
factchecking content can be most effective.
4 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?Introduction
A backfire effect, as originally conceived, is the idea that
when a factual claim aligns with someone’s ideological
beliefs and they are told that claim is wrong, they believe it
more strongly than if you hadn’t said anything. This type of
backfire is specifically a “worldview” backfire effect, and is also
sometimes referred to as a boomerang effect. 1
There is some evidence that this type of backfire effect may
occur in some cases—such as with particularly contentious
topics, or where the claim being asked about is ambiguous
or lacks clarity. 2 But there is a growing body of evidence that
in many cases it doesn’t. Recent studies have concluded that
“worldview backfire effects are not the norm and may occur
under very specific circumstances”, and that “citizens can
accept factual corrections of misstatements even when they
are made by one’s preferred candidate during a presidential
election”. 3
Much of the evidence we have on backfire comes from
laboratory and survey experiments, mostly in the United
States. The most recent studies now suggest that generally
debunks (see definitions box below) can make beliefs in specific
claims more accurate.
That said, we’ve still got a long way to go to understand why
beliefs in inaccurate claims persist and what can be done to
address them.
‘The most recent
studies now
suggest that
generally debunks
can make beliefs
in specific claims
more accurate.’
1 Other types of backfire include for example the “familiarity backfire effect” (the effect of repeating the disputed claim within a
debunk) where the evidence is similarly complicated. See: skepticalscience.com/Debunking-Handbook-Part-2-Familiarity-Backfire-
Effect.html and Swire B, Ecker UKH, Lewandowsky S, 2017. “The role of familiarity in correcting inaccurate information”, Journal of
Experimental Psychology: Learning, Memory, and Cognition, 43 (12), 1948-1961.
2 See Wood T, Porter E, 2017. “The elusive backfire effect: mass attitudes’ steadfast factual adherence”, forthcoming at Political
Behavior.
3 Swire B, Berinsky AJ, Lewandowsky S, Ecker UKH, 2017. “Processing political misinformation: comprehending the Trump
phenomenon”, Royal Society Open Science, 4 (3); Nyhan B, Porter E, Reifler J, Wood T, 2019. “Taking fact-checks literally but not
seriously? The effects of journalistic fact-checking on factual beliefs and candidate favorability”, forthcoming at Political Behavior.
5
Full Fact The backfire effect: does it exist? And does it matter for factcheckers?The research also shows that our attitudes can still affect the
extent to which a debunk might be effective—the likelihood
that someone will be persuaded by it. And there are questions
around how to ensure that debunks have lasting effects on
beliefs in claims.
Beyond experimental settings, we know that both in the UK
and worldwide there are specific inaccurate claims which are
still believed by many, despite widespread statements to the
contrary.
For example, Ipsos MORI’s 2017 Perils of Perception series
found about 20% of people across 38 nations believed it’s
true that there is a link between some vaccines and autism in
healthy children, and a further 38% said they didn’t know.
This is despite a large body of evidence 4 that, in the words of
the UK’s National Autistic Society, demonstrates “research has
comprehensively shown that there is no link between autism
and vaccines”. And there is evidence from the United States
that after President Obama published his birth certificate,
misperceptions about his nationality initially fell, but then in
time increased again.
This briefing primarily focuses on belief in the accuracy of
the specific claim in question. Some of the studies also look
at the extent to which debunks change people’s attitudes
or behaviour—to use the earlier example, you might be
persuaded that there is no link between vaccines and autism,
but at the same time become more convinced you don’t want
your child to be vaccinated. 5
These are also sometimes referred to as “backfire effects”,
but we leave them out of our definition of backfire. This is a
more complex area, and there is limited research addressing
how debunking changes underlying attitudes or behaviour.
For factcheckers there is also a question of how to interpret
the meaning of this research for our work, since factcheckers’
primary focus tends to be on promoting accurate information
and not on how individuals choose to make decisions on the
basis of that information.
4 See for example: Stratton K, Ford A, Rusch E, & Wright Clayton E, 2011. Adverse Effects of Vaccines: Evidence and Causality,
(Washington: National Academies Press); Taylor LE, Swerdfeger AL & Eslick GD, 2014, “Vaccines are not associated with autism: An
evidence-based meta-analysis of case-control and cohort studies”, Vaccine, 32 (29), 3623-9.
5 See for example: Nyhan B, Reifler J, Richey S, & Freed GL, 2014. “Effective messages in vaccine promotion: A randomized trial”,
Pediatrics, 133 (4), 835-842.
6
Full Fact The backfire effect: does it exist? And does it matter for factcheckers?Definitions
Factual claims: We use this to refer to statements about
the state of the world which Full Fact/factcheckers could
factcheck, as distinct from statements of opinion.
Corrections: The academic literature on this topic tends
to talk about statements which debunk other statements
as “corrections”. At Full Fact we talk about “corrections”
as public acknowledgements that factual claims were
inaccurate (such as a politician correcting the record,
or a newspaper correcting an article), so in this briefing
we’re just using the terms ‘debunking’ and ‘debunking
messages’, except where directly quoting from academic
studies.
Debunking: Factual messages which seek to rebut
inaccurate factual claims. Here we will use this as a
catch-all term, which includes debunks which might not
directly say a claim is wrong (for example they might
simply provide the accurate information, but not directly
call out the claim), or content where the rebuttal is only
one part of a broader journalistic report on the topic.
Factchecks: We will use this term specifically for
published articles in which the primary purpose of the
content is evaluating and explaining the accuracy of a
claim (such as might come from dedicated factchecking
organisations).
Attitude-inconsistent information: We use this for text
containing factual information that is inconsistent with
someone’s beliefs, but does not seek to directly address
specific factual claims quoted in the same place.
7
Full Fact The backfire effect: does it exist? And does it matter for factcheckers?How we’ve selected
the studies, and some
questions you might
consider
We have focused this briefing on the studies that we have seen
referenced the most by academics in relation to backfire effects 6 ,
and specifically on studies that examine the effects of debunks
of factual claims (rather than simply the effects of attitude-
inconsistent information). 7
We have also focused on experimental studies where
respondents are randomly assigned to treatment and control
groups.
We look in detail at seven studies, and reference the findings of
three further studies and three literature reviews. This briefing
explores the studies from a practical perspective. We’ve looked
at things like: what format of debunk has been tested (such as
within news articles or standalone conclusions)? How clear is the
debunk? What types of claims and claimants have been tested—
highly controversial ones, or less contentious topics? Who has
been sampled and how large is the sample?
Many of the studies tend to be conducted using participants
on Amazon’s Mechanical Turk (MTurk). MTurk is an online
marketplace where participants get paid for participating in
surveys. While its use is common in academic studies—it’s
cheaper than nationally representative sample providers—it has
some significant caveats to be aware of.
For one, people who choose to join the marketplace are known
to differ in at least some ways to those who don’t. In the United
States, Conservatives are known to be under-represented on the
platform for example. And secondly, people who are on MTurk
can fill out a lot of surveys, including on the same topic, and
there is evidence that at least some of them start to ‘game’ the
system according to what they think is expected or not expected
of them.
Some of the studies below have attempted to check
for differences, comparing MTurk samples to nationally
representative samples—and have generally found no significant
differences.
6 For example in Flynn DJ, Nyhan B, and Reifler J, 2017. “The nature and origins of misperceptions: understanding false and
unsupported beliefs about politics”, Political Psychology 38 (S1), 127-150); Lewandowsky S, Ecker UKH, Cook J, 2017. “Beyond
misinformation: understanding and coping with the “post-truth” era”, Journal of Applied Research in Memory and Cognition 6 (4), 353-
369; Nyhan B, Porter E, Reifler J, Wood T, 2019. “Taking fact-checks literally but not seriously? The effects of journalistic fact-checking
on factual beliefs and candidate favorability”, forthcoming at Political Behavior.
7 For this reason we have excluded Hart PS, Nisbet EC, 2011. “Boomerang effects in science communication: How motivated
reasoning and identity cues amplify opinion polarization about climate mitigation policies”, Communication Research, 39 (6), 701-
723, and Schaffner B and Roche C, 2016. “Misinformation and motivated reasoning: responses to economic news in a politicized
environment”, Public Opinion Quarterly 81 (1), 86-110.
8
Full Fact The backfire effect: does it exist? And does it matter for factcheckers?So, where has backfire
been observed?
There are two main studies that tend to be pointed to
as evidence of factual backfire taking place under some
circumstances (leaving out the studies on vaccine effects).
Both of these focus on short debunks as they might appear in
news articles, rather than full factcheck articles explaining the
analysis. They also test generally less conclusive debunking
messages compared to more assertive conclusions tested in
recent studies discussed below.
“When corrections fail: the persistence of
political misperceptions”
Brendan Nyhan and Jason Reifler (published in Political
Behavior, 2010)
The idea of a backfire effect first became particularly well
known from a study by Brendan Nyhan and Jason Reifler in
2010, where they found several instances of where “corrections
actually increase misperceptions among the group in
question”.
Their theory was that this happens because when people are
shown information that goes against their political beliefs, they
will “counterargue” it in their minds strongly enough that then
they’ll end up with stronger beliefs in the original inaccurate
claim than if they weren’t shown the debunking information.
Ever since, journalists and others have repeatedly referenced
this study (or related studies) to say that attempts to correct
people cause backfire effects. (One recent example, from
New York Times columnist Nicholas Kristof: “One challenge is
that fact-checking doesn’t work very well... This is called the
‘backfire effect.’”) However, even on this study the authors say
the effects were “overstated and oversold” and that they never
actually found that it always happened.
They conducted two studies, one in autumn 2005 and one in
spring 2006, testing five claims in total among samples of less
than 200 US undergraduates. They found evidence of backfire
in response to two claims they defined as inaccurate, both
made by ex-US President George W. Bush.
The first and most well-known claim, tested in the first study,
was about whether Iraq had weapons of mass destruction
(WMD) that were destroyed or hidden before US forces arrived
in the US invasion. Participants were shown a news article
about comments made by George Bush defending the decision
to invade Iraq and were then asked the extent to which they
9
Full Fact The backfire effect: does it exist? And does it matter for factcheckers?agreed with the statement “Immediately before the U.S.
invasion, Iraq had an active weapons of mass destruction
program, the ability to produce these weapons, and large
stockpiles of WMD, but Saddam Hussein was able to hide or
destroy these weapons right before U.S. forces arrived”.
Some had been shown the news article on its own, and some
saw the news article with a debunk within the article saying
that the Central Intelligence Agency had released a report to
say no weapons had been stockpiled. Conservatives who saw
the news article with the debunk ended up agreeing with the
statement more strongly than those who saw the news article
without the debunking element—a “backfire effect”.
The second claim displaying evidence of backfire, tested in the
second study, was about whether George Bush’s tax cuts paid
for themselves in increased government revenue.
There was no significant evidence of backfire in a further three
claims tested in the second study, including a simplified claim
about WMD. 8 There was some evidence of debunks having a
neutral effect.
So what can we make from this? The experiment purposefully
covered a highly controversial topic in American politics where
people would have prior beliefs, to differ from hypothetical
scenarios tested in previous research. So it’s arguably
unsurprising that individuals were unpersuaded by a single
news item. A recent study, which we’ll come on to, tested a
simplified version of the claim and found no backfire—which
could either reflect the lower intensity of the issue given it was
tested a few years later, or less ambiguity and nuance in the
wording of the claim.
The authors also said they purposefully tested debunks which
did not come down hard on what the truth was, so they would
appear like they do in news articles—in this instance, saying
that the Central Intelligence Agency had released a report to
say no weapons had been stockpiled, rather than explicitly
saying the claim was false. It’s possible this made the debunk
less convincing.
This study also had a small sample of fewer than 200 people—
which means it might have been a mistaken finding—and
was conducted among undergraduates rather than a national
sample.
More recent evidence has suggested that these findings may
have been due to the ambiguity of the statement tested, and
due to this smaller sample size. We’ll come back to this later.
8 This includes an additional claim mentioned in a footnote in the paper, where the authors say they tested corrections to a claim
made by Michael Moore (in the movie “Fahrenheit 9/11”) that the war in Afghanistan was motivated by oil company Unocal’s desire to
build a natural gas pipeline through the country. They found no relevant statistically significant effects.
10 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?“The hazards of correcting myths about
health care reform”
Brendan Nyhan, Jason Reifler and Peter Ubel (published in
Medical Care, 2012)
In this larger study of around 900 US participants from an opt-
in internet panel, Nyhan and Reifler worked with Peter Ubel to
test for backfire in response to a claim made by former Alaskan
Governor Sarah Palin. They found backfire in just one subgroup:
among Palin supporters with high political knowledge.
Participants were randomly assigned to either a control group
in which they read an article on Sarah Palin’s claims about
“death panels” or an intervention group in which the article
also contained corrective information refuting Palin. The claim
related to President Obama’s health reforms which Palin
claimed would create a “death panel” in which bureaucrats
decide whether to continue a person’s health care.
For people with high political knowledge and who supported
Sarah Palin, they found they clung more strongly to the beliefs
they held. This was the only group where backfire effects were
found in this study.
The study used a similar debunking style as the 2010 study
where they quoted someone else as saying the claim was
inaccurate, using the text: “However, non-partisan health care
experts have concluded that Palin is wrong”.
Brendan Nyhan and Jason Reifler have updated their beliefs
since these studies. Nyhan now says that facts can change
minds “in some cases”—the question is when and how durably
they can be changed. He points to macro public opinion data
which he says is less encouraging, as we can see from surveys
like Ipsos MORI’s Perils of Perception series.
Other studies that have found some form of
backfire
Some other studies have found evidence of backfire when it
comes to underlying behavioural intentions, but not for specific
belief in the claim debunked. For example, as referenced
earlier, two studies 9 testing the effects of debunks addressing
specific fears about vaccines found the debunks were
9 Nyhan B, Reifler J, Richey S, & Freed GL, 2014. “Effective messages in vaccine promotion: A randomized trial”, Pediatrics, 133 (4),
835-842; Nyhan B, Reifler J, 2015. “Does correcting myths about the flu vaccine work? An experimental evaluation of the effects of
corrective information”. Vaccine, 33(3), 459-64.
11 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?effective at reducing beliefs in the claim they were debunking,
but still decreased the intention to vaccinate among those with
high levels of concern about vaccine side effects.
This raises the question of what effect debunks or factchecks
can have (and should seek to have) on underlying attitudes
and behaviour—especially when the influence of our attitudes
on our factual beliefs is often hard to disentangle from the
extent to which our factual beliefs cause our attitudes. 10
10 Duffy B, 2018. The Perils of Perception: Why we’re wrong about nearly everything (London: Atlantic Books).
12 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?Recent attempts to
replicate have found no
backfire
Recent attempts to directly test for the backfire effect,
generally among much larger samples than previously used,
have found little evidence of the effect.
These studies have tested claims and debunks appearing on
their own, as well as claims and debunks appearing in a mock
news article.
“The elusive backfire effect: mass attitudes’
steadfast factual adherence”
Thomas Wood and Ethan Porter (published in Political
Behavior, 2018)
A recent study by Thomas Wood and Ethan Porter conducted
five studies testing 52 claims with more than 10,100 people in
the US.
They selected prominent issues, ranging from the incidence of
gun violence through to undocumented immigrants’ criminal
activities. They tested claims and debunks both as separate
pieces of text, and within mock news articles.
They conclude that backfire is “stubbornly difficult to induce”
and found: “The average subject exposed to the correction
subsequently expressed attitudes more in line with the facts.”
This includes with Nyhan and Reifler’s WMD text, where they
found no backfire.
That’s not to say there was no effect of partisanship—in the
majority of cases they still found that people were more or
less likely to believe a claim was true or false depending on
their political beliefs. But they found that debunking did move
individuals towards being more accurate on average, including
among both liberals and conservatives.
They say their results do “not lead us to conclude that
backfire is categorically impossible. Certain issues and certain
questions—perhaps asked at moment when ideology or
partisanship, or both, are particularly salient—might plausibly
trigger factual backfire”.
The types of debunks they tested varied from fairly non-
confrontational factual statements as might appear in news
articles such as “in fact, according to public records, homicides
13 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?of law enforcement officers have been declining for decades”,
to more direct debunks, for instance stating that a claim was
“plainly false”.
What makes their study particularly interesting is that they
also tested to see whether the complexity of the claims
participants were asked to agree or disagree with, and the
relevance of the debunk to the claims, made any difference to
the effect of the debunk.
Nyhan and Reifler’s original study asked respondents if it was
true that “Immediately before the U.S. invasion, Iraq had an
active weapons of mass destruction program, the ability to
produce these weapons, and large stockpiles of WMD, but
Saddam Hussein was able to hide or destroy these weapons
right before U.S. forces arrived.” To see what effect the
ambiguity of this statement had, Wood and Porter tested this
and a simpler version: “Following the US invasion of Iraq in
2003, US forces did not find weapons of mass destruction”.
They found no backfire from either version. However, the
simpler version was the only version of the question to
successfully elicit a change in responses from participants—
with liberals presented with the simpler version adopting the
debunk and conservatives showing no effect. Wood and Porter
suggest this might be because the original wording offered
more room for dispute, and was also just potentially too
complex for respondents.
When they varied the complexity of six other claims, they
found no level of complexity caused backfire, but more
complex statements did decrease the effect of the debunk
among conservative respondents.
Lastly, they asked respondents to rate the ‘relevance’ of
debunking statements to the original claim they were
debunking. Surprisingly, they found no significant differences
in the extent to which debunks changed people’s beliefs
depending on whether the debunk was seen as closely related
or tenuous.
All but the last of their studies were conducted with
participants from MTurk. Their last test compared results
between 1,000 Mechanical Turk participants and 1,000
participants from a nationally representative sample. They
found generally no significant differences in responses to
corrections, although people identifying in the centre of the
political spectrum on MTurk were found to be slightly more
responsive to new factual information than those in the
national sample.
14 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?“Taking fact-checks literally but not
seriously? The effects of journalistic fact-
checking on factual beliefs and candidate
favorability”
Brendan Nyhan, Ethan Porter, Jason Reifler and Thomas
Wood (forthcoming in Political Behavior, 2019)
Wood and Porter have since collaborated with Nyhan and
Reifler on a study at the height of the 2016 US Presidential
elections, and found even at the peak of election season
they could reduce misperceptions about trends in crime and
unemployment among both Republicans and Democrats,
based on claims made by Donald Trump.
They tested two claims: one made in Donald Trump’s
nomination acceptance speech that violent crime had
increased substantially, with the study conducted several
weeks after; and one made by Donald Trump in the first 2016
presidential debate that jobs were moving from the US to
Mexico, with the study conducted that evening. Misperceptions
decreased among both Clinton and Trump supporters,
compared to those who saw the claim without any corrections.
No Hillary Clinton claims were tested, so we only have evidence
of possible partisan reactions by Trump supporters.
The debunks they tested were similar to the fairly non-
confrontational debunks tested in previous studies and were
based on media debunks at the time. The first one started by
saying, “According to the FBI’s Bureau of Justice Statistics, the
violent crime rate has fallen dramatically and consistently over
time...”, and the second by saying, “In fact, according to the
Bureau of Labor Statistics, unemployment has fallen in both
states...”.
In the first study they tested to see what happens when you
include quotes from elites trying to undermine the debunking
information—in one instance where former Trump campaign
manager Paul Manafort questioned the validity of the statistics,
and in another where he suggested the FBI was biased to
the Clinton campaign. They found some evidence that these
reduced the effect of the debunk on Trump supporters.
They also asked respondents about how accurate they thought
the crime statistics were, and how fair and unbiased or biased
the article was. Despite their beliefs becoming more accurate,
Trump supporters were more likely to view the article as less
accurate and fair when there was a debunk. They were also
less likely to view the statistics as accurate when they were
used in a debunk, especially when the experiment said they
had been questioned by a Trump staffer.
15 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?Finally, both studies also tested to see if the debunk affected
citizens’ attitudes towards Trump and found it didn’t. This is
arguably unsurprising. As the authors comment, the study
only looks at the effect of a single debunk. Further research is
needed to see what the effect is on attitudes when individuals
see multiple debunks of claims by the same politician, and
when shown debunks of claims made by opposition politicians.
The first study was tested on around 3,000 Mechanical
Turk participants, and a further 1,200 from a nationally
representative sample. No significant differences were found
in the results—with both showing a significant decrease in
misperceptions among both Clinton and Trump supporters. The
second study was conducted with 1,500 MTurk participants.
“Processing political misinformation:
comprehending the Trump phenomenon”
Briony Swire, Adam Berinsky, Stephan Lewandowsky, and
Ullrich Ecker (published in Royal Society Open Science,
2017)
This study, a collaboration between Briony Swire, Adam
Berinsky, Stephan Lewandowsky and Ullrich Ecker, provides
further evidence of belief change after debunks. They tested
reactions to statements made by Donald Trump in the run-up
to the 2016 US Presidential election.
Around 1,800 US participants from MTurk took part in a first
study in November 2015, and around 1,000 US participants
from a national online sample took part in a second study in
July 2016.
They were primarily focused on how the source of the claim
influences the effectiveness of debunks. While we will not go
into these findings in detail, overall they found “no evidence for
a worldview backfire effect in either experiment”.
Participants were shown statements made by Donald Trump on
the campaign trail, half of which were accurate and the other
half inaccurate. Debunks/affirmations to the claims were then
presented, with a statement about the accuracy of the claim
(e.g. “This is false” or “This is true”) and an explanation of this
assessment with an explicit reference to a reputable non-
partisan source.
Unlike the other studies explored here, participants were asked
how accurate they thought the statement was before they saw
the debunk/affirmation, then given a reminder of their original
evaluation when they were asked again, either straight away
or after a week, about their belief in the statement. Despite
this they found both Democrats and Republicans showed a
“substantial amount” of belief change—although with some
decline after a week.
16 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?They found no change in participants’ voting preferences after
the debunks.
“Does truth matter to voters? The effects
of correcting political misinformation in an
Australian sample”
Michael Aird, Ullrich Ecker, Briony Swire, Adam Berinsky and
Stephan Lewandowsky (published in Royal Society Open
Science, 2018)
Building on the results of the previous research, this study
sought further evidence of whether debunks change factual
beliefs. Where the previous study used four accurate and four
inaccurate claims, this study looked at whether debunks of
disproportionate numbers of inaccurate and accurate claims by
a politician affect support for that politician.
It has a much smaller sample than the previous studies
discussed, with 370 Australian participants, including around
80 undergraduate students and the rest recruited online.
But it provides some useful preliminary evidence for possible
responses to debunks with a non-US sample.
Participants were shown short conclusions assessing the
accuracy of statements made by politicians from opposing
parties: Bill Shorten from the Labor party, and Malcolm Turnbull
from the Liberal party.
They found debunks of inaccurate claims reduced beliefs
to equally low levels across “left-wing” and “right-wing”
participants, and evaluations of accurate claims increased the
accuracy of beliefs across left-wing and right-wing participants.
They also found debunks of myths actually reduced beliefs
more strongly if the myths came from a favoured politician,
since participants were more likely to believe those myths in
the first place and therefore their beliefs had “further to fall”
(this was mainly among left-wing participants).
The authors said it’s possible these results were affected by the
claims covered in the study—which they said were unlikely to
have challenged strong beliefs, and which may have differed in
importance between the two politicians.
When altering whether participants were shown an equal
number of accurate and inaccurate claims by a politician,
or four false statements and only one true statement, they
found the latter caused significant decline in reported feelings
towards the politician being factchecked—at least within this
experimental setting.
17 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?The study reports the findings of a follow-up study conducted
in the USA, not published yet, which found Americans’ feelings
towards politicians barely shifted in the disproportionate
condition—unlike the Australian sample where there was a
significant reduction in beliefs. This highlights the importance
of carrying out more research on this topic across different
cultures and country contexts.
“Do people actually learn from fact-
checking? Evidence from a long-term
experiment during the 2014 campaign”
Brendan Nyhan and Jason Reifler (not peer reviewed)
The largest study of real-life factchecks is by Brendan Nyhan
and Jason Reifler, who looked at the effect of factchecks
produced by the US factcheckers PolitiFact in 2014.
In the study, which hasn’t gone through peer review, they
randomly exposed a representative panel of 1,000 Americans
to either factchecks or unrelated press releases over a series
of surveys during the 2014 mid-term election campaign. These
included true, false and half-true Democrat and Republican
claims.
They found factchecks significantly improved the accuracy
of people’s beliefs, with overall little difference by partisan
beliefs. This effect was strongest for those with high political
knowledge—with the proportion of correct answers increasing
by 17 percentage points from those who saw no factchecks
to those who saw the factchecks. For those with low political
knowledge, these proportions increased by 11 percentage
points.
Where the factchecks were very long, the study put together
shorter versions—including the introductory text, the final
conclusion and the truth rating. However, they provide the
results of a smaller experiment which tested the full-length
versions and found similar effects. (Those who saw the full-
length versions were more likely to say there was too much
detail.)
The authors say these findings are “relatively large”
considering participants were asked about the accuracy of the
claims a while after—participants completed their first survey
in September 2014 and then were shown the factchecks in a
series of three mini surveys between then and the final survey
in November 2014.
More research is needed to understand what might be
contributing to this. One possibility is this is due to participants
reading the factchecks unusually carefully, since the survey
tested if participants had read each factcheck and they had
18 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?to read the factcheck again if they answered a question about
it incorrectly. While their study didn’t compare factchecks to
the types of debunks embedded within news articles, they also
say that the findings could suggest the less ambiguous style of
factchecking may be more effective.
Other factors may have influenced it too. They say the topics
may have been less important than the controversial topics
explored in previous experiments. The factchecks shown to
participants also included claims from lesser-known state-
level politicians rather than national-level politicians, so
participants may have had fewer prior attitudes before reading
the factcheck.
Lastly, the factchecks tested included a truth rating, which
Full Fact doesn’t use. We see our job as filling in the shades of
grey when campaigners often talk in black and white. We don’t
know how much difference rating scales make—a separate
study, looking at the effect of truth scales using made-up
factchecks of fictional claims and politicians, found scales
made little difference for political claims, but we need more
evidence on this topic.
19 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?Other studies
Studies in the UK
There has also been one peer-reviewed study on this broad
topic conducted in the UK in 2013. This looks at the effect of
showing people information on the unemployment rate and
the growth rate on beliefs about recent economic performance.
They found providing this information generally made
individuals’ beliefs about recent economic performance more
accurate, regardless of their prior beliefs.
Studies looking at broader issues
This briefing has focused specifically on evidence relating to
the existence or not of backfire. There is a whole range of
further research still to explore on factors that improve the
receptivity of factchecks—for example, studies have suggested
that replacing inaccurate information with new information
is better than simply stating something is incorrect, and that
the formats of factchecks can make them more effective—
for example, videos and graphs have been found to be more
effective than text.
20 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?What does this all mean
for factcheckers?
This review suggests factchecking does help to inform citizens
and backfire effects are rare not the norm.
We still need more evidence to understand how factchecking
content can be most effective. A review of psychological
studies testing out methods to counter misinformation
concluded in 2017 that “Mounting evidence suggests that the
process of correcting misinformation is complex and remains
incompletely understood”. That is especially the case outside
the United States, which is where the majority of these studies
have been conducted.
The studies discussed here have shown that political beliefs
can affect the extent to which debunking can be successful. We
need more evidence to understand this in the United Kingdom
and in other countries outside the US. Evidence on the effect of
factchecks is promising, and needs exploring further as to how
we can make factchecks particularly effective.
‘We still need
more evidence to
understand how
factchecking content
can be most effective.’
One of the areas we are also interested in exploring more is
under what circumstances someone with deeply entrenched
inaccurate factual beliefs might be persuaded to change
them. For example, what difference does it make if the person
or organisation putting out the message is trusted by the
individual—and how might factchecking organisations gain
that trust? The paper by Briony Swire and others is one of a
number of papers building evidence on the former of these
questions.
Full Fact’s work aims to empower the public by holding those
in power to account for the information they use—to help
prevent inaccurate claims from being made and from being
spread before they become widespread misperceptions. We
do this by seeking corrections from politicians and journalists
to stop the spread of claims, and by seeking systemic changes
that improve the accuracy of public debate. We are working
to gather more evidence on the effectiveness of this in the
coming years.
21 Full Fact The backfire effect: does it exist? And does it matter for factcheckers?Full Fact
2 Carlton Gardens
London
SW1Y 5AA
team@fullfact.org
+44 (0)20 3397 5140
@FullFact
fullfact.org
[*/quote*]