Study: The Effect of watching CNN instead of Fox News

Mendel

Senior Member.
I'm posting this here because which facts people have access to changes what bunk they'll believe. It may help to steer conspiracy theorists towards more factual and less biased news sources, see https://www.metabunk.org/threads/media-bias.11554/ .

Article:
Fox News viewers who were paid to watch CNN for 30 days eventually became more skeptical and less likely to buy into fake news, according to a new study.

The study titled “The manifold effects of partisan media on viewers’ beliefs and attitudes: A field experiment with Fox News viewers” by David E Brockman and Joshua L Kalla was conducted in September 2020 and published last week.

“Of 763 qualifying participants, we then randomised 40 per cent to treatment group. To change the slant of their media diet, we offered treatment group participants $15 per hour to watch 7 hours of CNN per week, during Sept. 2020, prioritising the hours at which participants indicated they typically watched Fox News,” the study said.

The study found changes in attitudes and policy preferences about Covid-19, evaluations of then president Donald Trump and Republican candidates as well as elected officials.

“Despite regular Fox viewers being largely strong partisans, we found manifold effects of changing the slant of their media diets on their factual beliefs, attitudes, perceptions of issues’ importance, and overall political views,” the authors of the study said.

[...]

A survey was conducted after three days on the participants, which captured significant shifts in beliefs.

On Covid, viewers who switched were six percentage points more likely to believe other countries handled the virus better than the US, five points more likely to believe that people suffer from whats known as “long Covid”, and 11 points less likely to say a president should focus more on containing violent protests than the pandemic.

On Mr Trump, there was a nearly three point decrease in the “feeling thermometer rating” after switching from Fox to CNN.

On then-presidential candidate Mr Biden, switchers were 13 points less likely to agree that “we’ll see many more police get shot by Black Lives Matter activists” if he was elected and 10 points less likely to believe his supporters were happy when police officers get shot.

And on mail in voting, switchers were seven points more likely to support after watching CNN.

“We found large effects of watching CNN instead of Fox News on participants’ factual perceptions of current events (i.e., beliefs) and knowledge about the 2020 presidential candidates’ positions,” the researchers said. “They discovered changes in attitudes about Donald Trump and Republicans as well as a large effect on their opinions about Covid.”
 
I'd be curious to see this study across more than two news sources with only one direction of change. Is there a benefit to looking at and considering more points of view, or is it just a case of "if you look at news sources that I look at, you'll think more like me!"
 
if 3 hours of cnn can do all that, Metabunk must be doing something seriously wrong.
It's called expecting people to read. This is why Mick's videos are good, but producing videos that are short, well presented and accurate is a time consuming thing to do.
 
Maybe the reason the study’s participants views changed wasn’t due to the sense-making of CNN (since CNN makes little sense) but rather due to the fact that those who choose to watch Fox tend to be more willing to believe what they are told on tv.

To me, when a person gravitates towards extremes of any ideological spectrum, it’s often because their threshold for evidence is below average.
 
Last edited:
Maybe the reason the study’s participants views changed wasn’t due to the sense-making of CNN (since CNN makes little sense) but rather due to the fact that those who choose to watch Fox tend to be more willing to believe what they are told on tv.

To me, when a person gravitates towards extremes of any ideological spectrum, it’s often because their threshold for evidence is below average.
well the numbers aren't very big at all. 10% change would be like 30 people? (if those 30 people actually did exclusively watch Fox News..for 450$ in a month i would probably say i watch Hannity, even though i don't, and if my answers were different a month later it would mostly be because i forgot what i said the first time. :) )
 
I'd be curious to see this study across more than two news sources with only one direction of change. Is there a benefit to looking at and considering more points of view

Article:
If the full story of the Bidens’ international influence-peddling scheme had been told before the election, polls indicate it may have affected the result. Almost 50 percent of Biden voters knew nothing about Hunter’s laptop scandal, according to polling conducted after the election by the Media Research Center, and almost 10 percent said they would not have voted for Biden had they known.


MRC is right wingish, but since the liberal colleges won't take such a poll we're stuck with their data.

There is certainly a benefit of looking at more points of view because most media now-a-days is biased to highly-biased due to internet marketing issues. and all the big media channels/papers have been caught out lately telling whopper "mistakes".

jan 2020
Article:
These questions allowed us to measure things a few different ways. For example, we could examine trust gaps between different groups of people for different news outlets. We found that CNN is trusted by 70% of self-described liberal Democrats, but only 16% of conservative Republicans – a gap of 54 percentage points. Conversely, Fox News is trusted by 75% of conservative Republicans but only 12% of liberal Democrats – a 63-point gap.
 
Of 763 qualifying participants, we then randomised 40 per cent to treatment group.
Content from External Source
I like that someone watching Fox news needs "treatment".

What would this look like in reverse?

We assigned 40% of the MSNBC viewers to a treatment group and payed them $25.00 per hour (they demanded a living wage) to watch 7 hours of Fox News per week. Afterwards 8% changed from "Agree" to "Strongly Disagree" regarding the statement: "Hunter Biden never owned a laptop" and 7% changed from "Agree" to "Strongly Disagree" regarding the statement: "Hunter Biden was eminently qualified to sit on the bord of Barizma".;)
 
I like that someone watching Fox news needs "treatment".
That's just the language for this sort of study where you divide the subjects into a treatment group and a control group. It just means you're doing something to them you're not doing to the other group.

This is also why this study is notable. There've been numerous surveys linking media consumption (by type of media, and by outlet) to political opinion and factual knowledge, but those are correlations that hint at causations and don't prove them. This controlled study proves causation.
 
I'm not paying people to read posts and then take a quiz. I'm pretty sure we'd have a lot more effect (on those people) if we did.
we've had conspiracy theorists here longer than 3 hours, reading posts. i guess it's possible we've changed 10% of minds on general issues in 3 hours.


note: i cant access the article in OP and Mendel did not link the study, but i'm assuming that was a typo and should have said "30 days".
 
Article:
If the full story of the Bidens’ international influence-peddling scheme had been told before the election, polls indicate it may have affected the result. Almost 50 percent of Biden voters knew nothing about Hunter’s laptop scandal, according to polling conducted after the election by the Media Research Center, and almost 10 percent said they would not have voted for Biden had they known.
That's the NYP who published the "scandal" originally. There's still no proof that Joe Biden ever met the guy, and there's no "scheme".

Metabunk discussion of the laptop is at https://www.metabunk.org/threads/tucker-carlsons-missing-package.11432/ and https://www.metabunk.org/threads/hunters-laptop-etc.11427/ .
 
Fox News viewers who were paid to watch CNN for 30 days [...] we offered treatment group participants $15 per hour to watch 7 hours of CNN per week, during Sept. 2020, [...] A survey was conducted after three days on the participants,
if 3 hours of cnn can do all that, Metabunk must be doing something seriously wrong.
note: i cant access the article in OP and Mendel did not link the study, but i'm assuming that was a typo and should have said "30 days".
It's clear from the article that the treatment period lasted 30 days. I read the article to mean that the survey was administered 3 days after the treatment period ended. You can find similar articles (and from the looks of it, reprints) if you do a websearch on the title of the study.

The study itself is available at https://osf.io/jrw26/ .
Starting three days after the incentivized viewership period ended, 744 of the 763 study participants answered a follow-up survey.
Content from External Source
 
Last edited:
I'd also suggest that attaching a monetary value to one set of opinions might endow them with added gravitas. In many ways this study feels like having been (consciously or unconsciously) aimed at a desired conclusion.
 
(if those 30 people actually did exclusively watch Fox News..for 450$ in a month i would probably say i watch Hannity, even though i don't, and if my answers were different a month later it would mostly be because i forgot what i said the first time. :) )
1) The study isn't about people watching CNN "exclusively", they can still watch as much Fox as they want. It's just probably an hour less than before.

2) The participants actually watched CNN.
We enforced compliance with viewership quizzes (e.g., about which guest had just appeared), described in more detail below. Compliance with watching CNN was high in the treatment group; the median participant correctly answered 14 of the 15 quiz questions we asked over the course of the five quizzes.
Content from External Source
3) The study design compensates for these effects, because the control group gets surveyed the same, would also "forget what they said", etc. For your assumption, the observed change should be random, and not systematically in the same direction.
 
Last edited:
I'd also suggest that attaching a monetary value to one set of opinions might endow them with added gravitas. In many ways this study feels like having been (consciously or unconsciously) aimed at a desired conclusion.
How?
When the survey is held, the study is over. No respondent gains anything by answering one way or the other.

Do you feel the participants took CNN more seriously than they take Fox? Or would it be the same at best?

The conclusion is that watching CNN while getting paid changed people. Your argument is that watching CNN for a month while not getting paid would have had a significantly different outcome?
 
How is this study’s methodology distinguishable from indoctrination?

(Edited for brevity)
 
Last edited:
1) The study isn't about people watching CNN "exclusively", they can still watch as much Fox as they want. It's just probably an hour less than before.

2) The participants actually watched CNN.
We enforced compliance with viewership quizzes (e.g., about which guest had just appeared), described in more detail below. Compliance with watching CNN was high in the treatment group; the median participant correctly answered 14 of the 15 quiz questions we asked over the course of the five quizzes.
Content from External Source
3) The study design compensates for these effects, because the control group gets surveyed the same, would also "forget what they said", etc. For your assumption, the observed change should be random, and not systematically in the same direction.
By sheer coincidence, just hours before that study was released, I'd been trying to
convince a friend that folks that she wrote off as "unreachable" weren't always.
That, while I think a brief discussion with someone rarely changes minds, my
personal experience was that relatively adamant folks would sometimes budge
if given gradual, small bites of fact, that were being ignored or played down by
the news sources they were locked into.

I see why some would reflexively see this as a left/right thing...but I didn't...
I saw it more as a tunnel vs a tunnel with a few new vistas.
 
How is this study’s methodology distinguishable from indoctrination?

(Edited for brevity)

If it was attempted, which it wasn't, then it didn't work:
"We also found no evidence of impacts on outcomes measuring topics not directly covered on either network, such as support for democratic norms"
The subjects went in conservative/republicans, and came out conservative/republicans: "we did not find, nor did we expect to find, any evidence of long-run effects on candidate choices". They just have a slightly lower opinion of Fox News afterwards.
 
How is this study’s methodology distinguishable from indoctrination?

(Edited for brevity)
How is this study's methodology like indoctrination??

the process of repeating an idea or belief to someone until they accept it without criticism or question
Content from External Source
 
The study itself is available at https://osf.io/jrw26/ .
Starting three days after the incentivized viewership period ended, 744 of the 763 study participants answered a follow-up survey.
see that makes more sense, hence my joke at the top of the thread noone "got" or even thought about.

1) The study isn't about people watching CNN "exclusively", they can still watch as much Fox as they want. It's just probably an hour less than before.
i did not say they did. The point is that you can change your mind without watching CNN. So without having access to the study, i would assume they only hired people who only consumed Fox News and no other media. You might need to reread my comment.



10% is huge when you need to swing an election.
i know. that's why i quoted :
and almost 10 percent said they would not have voted for Biden had they known.


but in this study it is only 30 people. (or 45-15 assuming a 5% margin of error). And to think 10% of Republicans would change their vote because they watched an hour of liberal media a day, makes no sense. We're surrounded by liberal media in America, and we still vote Republican.

I'm glad that you're excited about your little study, but i wouldn't get your hopes up that 1 hour of liberal media is gonna make that much of a difference in elections. You haven't seemed to change your mind about anything political, and since you keep teaching people about the importance of consuming media with different points-of-view, i assume you have been periodically consuming right leaning media.
 
I see why some would reflexively see this as a left/right thing...but I didn't...
I saw it more as a tunnel vs a tunnel with a few new vistas.
i think people see it that way because they didn't ask MSNBC people to watch Fox News. But i agree with you, that i saw it as a liberal biased study that did highlight the need for all people to consume news from multiple media organizations.

Of course, we could be teaching this kind of thing in grade school. ie. how to parse information and look for contradictory information, how to recognize bias in media etc. It's not rocket science; you are what you eat.

But i dont see how policy makers can address some of these challenges. I would like the Fairness doctrine to be reinstated IF a show is under the label "News" or a person calls themselves a "journalist", but that isnt going to help with organizations choosing not to report on something.


Second, our results indicate challenges that partisan media may pose for democratic account-
ability. Our findings suggest that partisan media may affect voters’ choices at least in part because
it hides information about aligned incumbents’ failures and distorts perceptions of political rivals.
This suggests that partisan media does not only present a challenge for the opposing party, it may
present a challenge for democracy which may deserve attention from policymakers.

We also conducted an endline survey several weeks later that found these impacts largely re-
ceded as treated participants primarily returned to their prior viewership habits. Our finding that
participants’ attitudes meaningfully shifted at first away from and then back towards their partisan
side along with changes in their viewership behavior
Content from External Source
https://osf.io/jrw26/
 
How is this study's methodology like indoctrination??

the process of repeating an idea or belief to someone until they accept it without criticism or question
Content from External Source
I don’t see how the study determined that it was critical thinking that led some participants to adopt CNN’s views, rather than brute influence.

If a large group of apolitical people were paid to listen to Ben Shapiro or Rachel Maddow or Luis Elizondo or Fred Rogers for 7 hours a week, I’d imagine a small percentage of them might start to adopt some of their beliefs too. I’d say that would resemble indoctrination more than an increase in their ability to think critically.

How did the study differentiate between the people who were better informed and the people who were just parroting what they’d been repeatedly exposed to?
 
If it was attempted, which it wasn't, then it didn't work:
"We also found no evidence of impacts on outcomes measuring topics not directly covered on either network, such as support for democratic norms"
The subjects went in conservative/republicans, and came out conservative/republicans: "we did not find, nor did we expect to find, any evidence of long-run effects on candidate choices". They just have a slightly lower opinion of Fox News afterwards.
Thanks. I couldn’t access the full text.

Maybe I wasn’t clear. I don’t claim the study was an “attempt” at indoctrination (that would be really weird), only that the process of repeatedly exposing someone to a particular point of view resembles indoctrination.

CNN doesn’t promote democratic party membership, so I’m not sure how republicans remaining republicans through the study proves that the participants new perspectives on some issues were the result of CNN sense-making and not just repeated exposure to CNN perspectives.
 
If a large group of apolitical people were paid to listen to Ben Shapiro or Rachel Maddow or Luis Elizondo or Fred Rogers for 7 hours a week, I’d imagine a small percentage of them might start to adopt some of their beliefs too. I’d say that would resemble indoctrination more than an increase in their ability to think critically.

How did the study differentiate between the people who were better informed and the people who were just parroting what they’d been repeatedly exposed to?

while i do think watching/reading alternative media sources is essential for critical thinking, you do have a point.

Skimming the paper it seems the treatment group were given quizzes often to determine if they actually watched the CNN shows. so you are kinda conditioned to answer quiz questions on what the shows were about (vs. your opinion on them). I think this too could have skewed the final survey results a bit if readers saw it more as a "test".

it depends on the question though, for Ex:

five points more likely to believe that people suffer from whats known as “long Covid”, and 11 points less likely to say a president should focus more on containing violent protests than the pandemic.

some people suffer from long covid is a fact. it's good people learned this if they didn't know it originally, but the wording "likely to believe" indicates to me that they knew it before and they just didn't believe it. but after a month of fear mongering* about covid , as seen in part 2 of the quote, they were 'brainwashed/scared' into "believing it".



*i was pro-fear mongering over the virus, so i'm ok with that kind of brainwashing. although i disagree with the premise of part 2, both violence and covid are equally important to address. and they entail different parts of the government, so why would 1 have to have priority over the other?


@deirdre, @Mendel, please stop arguing about swinging elections, that's not what this paper is about.
of course it has to do with improving critical thinking skills when voting. otherwise what is the point? if all people were well informed on all the issues, our country wouldn't be so divided and maybe something constructive would actually get done once in a while to tackle our issues that hurt the population.
 
while i do think watching/reading alternative media sources is essential for critical thinking…
I absolutely agree that watching/reading alternative media sources is essential.

The person who listens to Ben Shapiro and Rachel Maddow and Luis Elizondo and Fred Rogers (which I literally do) is better informed than someone who listens to just one of them.

Three out of the four frustrate me. (Fred is usually spot on) And when they frustrate me it’s because they’re deaf to valid points made by the opposition.

Why this study did not seek to replicate its results by running it back with CNN viewers paid to watch Fox is baffling to me.

My naive perspective is:
Conservatism isn’t right. Liberalism isn’t right. Both are the graduations of a dial that requires frequent and careful adjustments.
 
Last edited:
How?
When the survey is held, the study is over. No respondent gains anything by answering one way or the other.

Do you feel the participants took CNN more seriously than they take Fox? Or would it be the same at best?

The conclusion is that watching CNN while getting paid changed people. Your argument is that watching CNN for a month while not getting paid would have had a significantly different outcome?
My argument is that by setting up a study the looked at user of one news provider, showing them only one other news provider (and not, for example, paying a group of CNN viewers to watch Fox), and attaching monetary value to the news provider you like, you can't be super surprised at the outcome -- the outcome is likely shaped in part by the way the study was designed.
 
Why this study did not seek to replicate its results by running it back with CNN viewers paid to watch Fox is baffling to me.
maybe they couldnt predict what Fox news would be talking about in September, so they couldnt write the "before" questions.
 
I'm posting this here because which facts people have access to changes what bunk they'll believe. It may help [the CTists to escape the rabbit hole] to steer conspiracy theorists towards more factual and less biased news sources, see https://www.metabunk.org/threads/media-bias.11554/ .
Article:
Media-Bias-Chart-9.0_Jan-2022-Unlicensed-Social-Media_Hi_Res-scaled.jpg

I tried using the dynamic chart and set it to TV only, but my tablet wouldn't have it. I wonder whether the researchers considered using npr instead of CNN.

I'd look for "indoctrination" in the lower half of the chart.
 
Last edited:
I'd look for "indoctrination" in the lower half of the chart.
it depends on what shows specifically you are watching. the shows on all those channels vary considerably. That's why it says "high variation in reliability" on your chart.

I wonder whether the researchers considered using npr instead of CNN.
they weren't looking for a more neutral or trustworthy station.
page 31 ...as these pre-existing liberal attitudes were primed by CNN’s agenda-setting efforts.

[pge 36] Third, since our intervention held constant the presence of partisan media, and simply shifted its slant (from Fox News to CNN), our estimates do not capture any effects of simply watching media.


....
While our theoretical argument would expect similar effects among viewers of other partisan media networks, future work should attempt to replicate this (e.g., shift MSNBC viewers to Fox News)
Content from External Source
page 2
The coverage of COVID-19 it did
offer provided little of the information CNN did, instead giving viewers information about why
the virus was not a serious threat.

On the other hand, Fox News extensively but highly selectively covered racial issues, and its coverage of these issues provided extensive information about Biden and other Democrats’ supposed positions on them and about outbreaks of violence at protests for racial justice in American cities. CNN provided little information about either.

The networks both covered the issue of voting by mail, but again covered dramatically different information about it
(in addition to offering different frames).
How would viewing networks with such different content affect viewers’ beliefs and attitudes?
Content from External Source
https://osf.io/jrw26/
 
note: at 9 and 10 oclock were the Chris Cuomo shows and Don Lemon shows on CNN. These people were watching pundit shows, not just regular Fox News or CNN news shows that air earlier.


7
The CNN line-up during these hours (the shows we drove participants
to) was Erin Burnett OutFront, Anderson
Cooper 360, Cuomo Prime Time, and the first hour of CNN Tonight with Don Lemon.

The Fox News line-up during
these hours (the shows we pulled participants away
from) was The Story with Martha MacCallum, Tucker Carlson
Tonight, Hannity, and The Ingraham Angle.
Content from External Source
each station gets progressively worse as the night gets later.


add: i say "pundit", but an article on Jen Psaki joining MSNBC called these shows "perspective programming". programming as in "a show" is called a program, not like cult programming...although that probably works for the late night shows on all these channels if you ask me.
 
Last edited:
Corroborating claims and content is worlds more beneficial than just ditching entire sources, important to keep in mind. For truly good practice ditching is saved for outlets or personalities you can confidently assess solely or mostly push out misleading or false material.
 
Corroborating claims and content is worlds more beneficial than just ditching entire sources, important to keep in mind. For truly good practice ditching is saved for outlets or personalities you can confidently assess solely or mostly push out misleading or false material.
I disagree. Mixing actual facts in with the misinformation is no excuse. Regularly pushing false material, and failure to announce corrections when they do, should be ample cause to drop a source.

It's established that false information, even if proven false, lodges in our minds, and can influence our thinking. I strongly advise stopping to habitually expose yourself to sources that unabashedly push proven misinformation. You'll be manipulated if you don't.

Corroborating claims and content is useful even when they come from reliable sources; it leads to a fuller, more nuanced picture, and is an important element of basic online literacy. See my post here for more details: https://www.metabunk.org/threads/how-to-detect-a-biased-source.11955/post-255978
 
Last edited:
That is what he said:
I know what he said.

@WhistlingWinds criterium is "solely or mostly" push out misinfo. In simplified language, he would "ditch" someone who lies more than they tell the truth.

My criterium is "regularly" and "unabashedly", i.e. with no correction, publish misinformation. I would ditch someone who always mixes 3 facts and 1 lie, while WW would not.
It's simply too much work for me to always have to question the facts. I could never trust anything they say.
I like sources who habitually get their facts straight, because then all I need to account for is bias, which is easier.
 
Back
Top