Fox News and Error Checking (from Wisconsin thread)

[jcow79]

Didn't the host correct himself later on the program?

It was corrected later on, yes. But that wasn't my point. I was questioning Dalton's suggestion that it may have been a simple graphic mistake. My assumption is that the things that make it to air have had some level of discussion or context prior. So at some point the results of this poll had to come in, some people had to discuss what it all meant and how they would present it on air, but the result was that the graphic and the script were BOTH wrong. To me this would exclude a graphic error exclusively unless both the graphic guy and the script guy made the same typo. This doesn't preclude an honest mistake somewhere up the production line but a much larger mistake regarding the entire context and results of the poll. (**Note: Composed prior to Dalton's reply so I'll address that next.)

It's from the same source. That segment was written by a producer, who ordered a graphic to go with the story. If the producer had bad data, it would be disseminated to both the host via the script (who more or less goes by the prompter) and the designer (who does what they're told).

So this would make it a mistake further up the chain rather than a graphics error which to me escalates this from simple common mistake to a producer or other information gatherer with severe reading comprehension issues or dishonest intent. I mean the results of this poll set the entire context of the discussion that would follow from it, why would no one catch that big of an error?
Certainly honest mistakes are bound to happen, you yourself seem to attest that mistakes can and do occur, but Fox admitted to the world they lie and distort the truth and it's their right to do so. Why shouldn't anything even slightly resembling that be considered as such unless proven otherwise?

[Dalton]

Not implausible, just not supported. You have to remember though that as someone in the trenches I have a difficult time accepting your confirmation bias theory, just based on personal experience.

Could you expand on why your personal experience leads you to think I am mistaken?

Because errors are easy to miss - not because they don't jive with the network's agenda, but because people are either too busy, too distracted or too ignorant to notice. You assume that every person working in the control room knows the exact details of every graphic, including whether or not it's correct, but this simply isn't the case. Do you know how many times people are misnamed because the operator doesn't know who the person on the screen is? Do you know how many errors are made because people are too distracted to double-check the fine details? Take that poll graphic, for instance - there are several people who would theoretically see the finished graphic before air. The designer doesn't know any better because they just did what they were told. The graphics playout operator doesn't know any better, because they don't have the details in front of them. The producer likely doesn't know any better, because they either screwed up the numbers in their own head (it happens, we're human) and didn't check the graphic thoroughly before verifying the order. This isn't confirmation bias - this is live news. Your theory relies too much on "Welp, looks good to me, send it on through!" and not enough "I can't be bothered to fact-check this fucking thing since I have TWENTY OTHER GRAPHICS in this block alone to worry about".

It's from the same source. That segment was written by a producer, who ordered a graphic to go with the story. If the producer had bad data, it would be disseminated to both the host via the script (who more or less goes by the prompter) and the designer (who does what they're told).

So this would make it a mistake further up the chain rather than a graphics error which to me escalates this from simple common mistake to a producer or other information gatherer with severe reading comprehension issues or dishonest intent. I mean the results of this poll set the entire context of the discussion that would follow from it, why would no one catch that big of an error?
Certainly honest mistakes are bound to happen, you yourself seem to attest that mistakes can and do occur, but Fox admitted to the world they lie and distort the truth and it's their right to do so. Why shouldn't anything even slightly resembling that be considered as such unless proven otherwise?

First, I would like a source on that Fox admission. Second, you provide a false dilemma here: either the producer is stupid or dishonest. However, you forget the many, many other potential failure points in these kinds of scenarios, probably because you don't understand the nature of live news. Was the source wrong? Did the producer transpose the numbers? Did something get lost in the shuffle? Did the producer realize they made a mistake? Do you realize how insanely busy a newsroom gets during the day? When you have an hour to fill and multiple blocks of stories for that hour, it is very, very easy to lose track of things. I am willing to give that producer the benefit of the doubt, because if that error was deliberate, that producer would likely be reprimanded. Even at Fox News.

[jcow79]

First, I would like a source on that Fox admission.

Jane Akre and Steve Wilson, two journalists of WTVT
Fox was able to appeal a lower court that awarded them damages based on whistleblower protection violations. Fox's defense:

successfully argued that the FCC policy against falsification was not a "law, rule, or regulation", and so the whistle-blower law did not qualify as the required "law, rule, or regulation" under section 448.102.

(quoted from the linked articles)
Although this defense was clever, it came with consequences. I hope the money they saved from the lawsuit was worth destroying their organizations credibility.

Second, you provide a false dilemma here: either the producer is stupid or dishonest.

True, admittedly the "stupid" assertion was a jab towards Fox producers, but my overall statements indicated I'm not unwilling to extend other considerations that could lead to an error like this, but I feel that Fox News is undeserving of them because of their their patterns of similar behavior and their legal defense regarding falsification.

However, you forget the many, many other potential failure points in these kinds of scenarios, probably because you don't understand the nature of live news. Was the source wrong? Did the producer transpose the numbers? Did something get lost in the shuffle? Did the producer realize they made a mistake? Do you realize how insanely busy a newsroom gets during the day? When you have an hour to fill and multiple blocks of stories for that hour, it is very, very easy to lose track of things. I am willing to give that producer the benefit of the doubt, because if that error was deliberate, that producer would likely be reprimanded. Even at Fox News.

I am not willing to give them that benefit. I feel the way they have conducted themselves in the past and routinely continue to conduct their broadcasts shifts the burden of proof to them. If their "mistakes" conveniently align with their broader narrative there's little reason not to consider they were intentionally caused, or intentionally overlooked.

[Simon_Jester]

Because errors are easy to miss - not because they don't jive with the network's agenda, but because people are either too busy, too distracted or too ignorant to notice. You assume that every person working in the control room knows the exact details of every graphic...

Um, no I don't?

My point is that if there are only a few people, or one person, who are in a position to check a fact before it gets on the air... it matters if that person is biased. Remember my "Larry Moe and Curly" example. That assumes three people have at least a 4 in 5 chance of spotting the mistake before it airs, each. Even that may be too generous, I know.

It doesn't matter. Whatever the percentage of errors is, when a significant chunk of your potential error-spotters* are biased, your error-checking will go to shit whenever the error supports those biases. Again, think about the "Larry Moe Curly" situation: only Curly has political bias, and even so it would be very easy for the rate of errors that support Curly's private personal bias to double compared to the baseline rate. All that would need to happen is that he'd have to go from spotting four of five errors to spotting three of five.

You don't need an army of people all checking each fact on the graphic, or in a position to check each fact on the graphic, for this to happen.

*(whichever lucky few do have the time and energy to spot their mistakes)

This isn't confirmation bias - this is live news. Your theory relies too much on "Welp, looks good to me, send it on through!" and not enough "I can't be bothered to fact-check this fucking thing since I have TWENTY OTHER GRAPHICS in this block alone to worry about".

Remember Larry, Moe, and Curly? That graphic may pass through the hands of a dozen other people, none of whom have time to do fact-checking. It doesn't matter; it all comes back to the same people who do error-check.

I'm damned sure that someone checks the stuff on TV, because you guys get too much stuff right. Maybe the checking is all done by whoever's responsible for taking note of a fact, one guy per fact, one potential point of error-checking.

That doesn't change the situation. You still get a disproportionate number of errors sliding through gaps in people's radar when those errors promote a popular bias within the organization. And acknowledging that very few people are in a position to fact-check content in TV news will not make this less likely.

First, I would like a source on that Fox admission. Second, you provide a false dilemma here: either the producer is stupid or dishonest. However, you forget the many, many other potential failure points in these kinds of scenarios, probably because you don't understand the nature of live news. Was the source wrong? Did the producer transpose the numbers? Did something get lost in the shuffle? Did the producer realize they made a mistake? Do you realize how insanely busy a newsroom gets during the day? When you have an hour to fill and multiple blocks of stories for that hour, it is very, very easy to lose track of things. I am willing to give that producer the benefit of the doubt, because if that error was deliberate, that producer would likely be reprimanded. Even at Fox News.

Dalton? Please, chill. This is not a slur on your profession, and I for one do understand that of the hands through which each individual bit of information on a news show passes, the vast majority are in no position to catch mistakes.

It doesn't affect my argument, which is a statistical one: if there is even one person who is in a position to spot errors, and that person is biased, or if there are multiple people in such a position and even one of them is biased... the rate of errors that favor the bias goes up. If the bias is strong or widespread, the rate of errors that favors the bias will skyrocket.

And that's before we even consider the possibility of any kind of intentional falsification, which I am leaving out of this whole question entirely.

OK?

[Dalton]

First, I would like a source on that Fox admission.

Jane Akre and Steve Wilson, two journalists of WTVT
Fox was able to appeal a lower court that awarded them damages based on whistleblower protection violations. Fox's defense:

successfully argued that the FCC policy against falsification was not a "law, rule, or regulation", and so the whistle-blower law did not qualify as the required "law, rule, or regulation" under section 448.102.

(quoted from the linked articles)
Although this defense was clever, it came with consequences. I hope the money they saved from the lawsuit was worth destroying their organizations credibility.

You claim that Fox admitted to "lie and distort the truth" and that it's "their right to do so". That's a wild exaggeration of the above and looks more like your specific takeaway from the incident.

Second, you provide a false dilemma here: either the producer is stupid or dishonest.

True, admittedly the "stupid" assertion was a jab towards Fox producers, but my overall statements indicated I'm not unwilling to extend other considerations that could lead to an error like this, but I feel that Fox News is undeserving of them because of their their patterns of similar behavior and their legal defense regarding falsification.

I am not willing to give them that benefit. I feel the way they have conducted themselves in the past and routinely continue to conduct their broadcasts shifts the burden of proof to them. If their "mistakes" conveniently align with their broader narrative there's little reason not to consider they were intentionally caused, or intentionally overlooked.

That makes you automatically biased in your evaluation, since you are no longer willing to consider other factors simply because of the organization that made the error. There is always doubt.

Because errors are easy to miss - not because they don't jive with the network's agenda, but because people are either too busy, too distracted or too ignorant to notice. You assume that every person working in the control room knows the exact details of every graphic...

Um, no I don't?

I retract my statement.

My point is that if there are only a few people, or one person, who are in a position to check a fact before it gets on the air... it matters if that person is biased. Remember my "Larry Moe and Curly" example. That assumes three people have at least a 4 in 5 chance of spotting the mistake before it airs, each. Even that may be too generous, I know.

It doesn't matter. Whatever the percentage of errors is, when a significant chunk of your potential error-spotters* are biased, your error-checking will go to shit whenever the error supports those biases. Again, think about the "Larry Moe Curly" situation: only Curly has political bias, and even so it would be very easy for the rate of errors that support Curly's private personal bias to double compared to the baseline rate. All that would need to happen is that he'd have to go from spotting four of five errors to spotting three of five.

You don't need an army of people all checking each fact on the graphic, or in a position to check each fact on the graphic, for this to happen.

*(whichever lucky few do have the time and energy to spot their mistakes)

Assuming, of course, that the people doing the error-checking are not ignorant of the facts. A misspelling is far easier to notice than transposed poll numbers (and even then people sometimes simply don't catch it - case in point, as of this writing we had a lower third banner up with a misspelled word for the better part of a minute and nobody but me noticed). Someone may think "that doesn't look right" and send if off anyway, since they don't know for sure it's wrong and don't have time to check.

Remember Larry, Moe, and Curly? That graphic may pass through the hands of a dozen other people, none of whom have time to do fact-checking. It doesn't matter; it all comes back to the same people who do error-check.

I'm damned sure that someone checks the stuff on TV, because you guys get too much stuff right. Maybe the checking is all done by whoever's responsible for taking note of a fact, one guy per fact, one potential point of error-checking.

That doesn't change the situation. You still get a disproportionate number of errors sliding through gaps in people's radar when those errors promote a popular bias within the organization. And acknowledging that very few people are in a position to fact-check content in TV news will not make this less likely.

Let's clear something up here. There's fact-checking, and there's error-checking. Fact-checking happens before and after the show, when it's easier to corroborate the data with the source. Error-checking is far more difficult, especially when you're rushing to get the graphic on air and have no time to do the fine proof-reading, especially if you are ignorant of the data. If someone is aware of the truth and fails to notice an error because of that, fine, that's confirmation bias, but your average control-room operator isn't always aware of what's correct and what's not correct.

This leaves open the possibility that the fact-checker - the producer writing the story - is either deliberately falsifying data or has a confirmation bias. I will admit that that's a greater possibility further up the line.

Dalton? Please, chill. This is not a slur on your profession, and I for one do understand that of the hands through which each individual bit of information on a news show passes, the vast majority are in no position to catch mistakes.

It doesn't affect my argument, which is a statistical one: if there is even one person who is in a position to spot errors, and that person is biased, or if there are multiple people in such a position and even one of them is biased... the rate of errors that favor the bias goes up. If the bias is strong or widespread, the rate of errors that favors the bias will skyrocket.

And that's before we even consider the possibility of any kind of intentional falsification, which I am leaving out of this whole question entirely.

OK?

Do I seem agitated? I'm just having a discussion. The main reason why I don't buy your confirmation bias theory is because a person has to be in the position of potentially knowing there's an error in the first place. This simply isn't the case. I will admit that in some cases errors might slip through with the person full-well knowing, if they noticed, that it was wrong - but in many cases nobody will realize, even for a long time, that there was a mistake.

Deliberate falsification is another story. There are dozens of unscrupulous people on both sides that I am sure feed biased data down the line. However, there is always room for doubt.

[Simon_Jester]

My point is that if there are only a few people, or one person, who are in a position to check a fact before it gets on the air... it matters if that person is biased. Remember my "Larry Moe and Curly" example. That assumes three people have at least a 4 in 5 chance of spotting the mistake before it airs, each. Even that may be too generous, I know.

It doesn't matter. Whatever the percentage of errors is, when a significant chunk of your potential error-spotters* are biased, your error-checking will go to shit whenever the error supports those biases. Again, think about the "Larry Moe Curly" situation: only Curly has political bias, and even so it would be very easy for the rate of errors that support Curly's private personal bias to double compared to the baseline rate. All that would need to happen is that he'd have to go from spotting four of five errors to spotting three of five.

You don't need an army of people all checking each fact on the graphic, or in a position to check each fact on the graphic, for this to happen.

*(whichever lucky few do have the time and energy to spot their mistakes)

Assuming, of course, that the people doing the error-checking are not ignorant of the facts. A misspelling is far easier to notice than transposed poll numbers (and even then people sometimes simply don't catch it - case in point, as of this writing we had a lower third banner up with a misspelled word for the better part of a minute and nobody but me noticed). Someone may think "that doesn't look right" and send if off anyway, since they don't know for sure it's wrong and don't have time to check.

Dalton, if the people doing the error-checking cannot detect a given error, then they don't count as error-checkers for the purpose of that error.

But it still doesn't change much of anything, you see. Suppose that when we are looking at transposed poll numbers, there are only two people who are in any position to check the numbers- say, Larry and Curly, who are both directly involved in doing the Internet research or whatever to look up the poll numbers so that the rest of the team can put them on the script/graphics/et cetera. Moe can check spelling errors, but not 'transposed number' errors, so he isn't counted.

Nothing changes. If Curly is still biased as before, you will still see a disproportionate number of errors favoring causes that Curly is biased in favor of.

All else being equal, errors would be expected to show no ideological slant. (D) will be accidentally written as (R) and (R) will be accidentally written as (D) with equal frequency. Poll numbers will be transposed to exaggerate the number of righties as often as they exaggerate the number of lefties. And so on. Because the errors occur randomly- rain falling on the just and unjust alike, as it were.

But when there is a nonrandom factor influencing the team's ability to spot errors, that factor will show up in the frequency of various kinds of errors. The errors are not created by any kind of malice, conspiracy, or bias. The ability to detect and correct the errors is not intentionally affected by any kind of malice, conspiracy, or bias. But the unintended interaction between the biased people on the staff and the team's ability to spot mistakes will still be there, as long as anyone on the team is ever in a position to catch the mistake- including the guy who looked up the poll numbers in the first place.

This half the reason why objectivity and neutrality and freedom from bias are so important in science, statistics, journalism, and so on. It's not just that un-objective people will deliberately twist the facts to their advantage. It's that they're less likely to notice when they (or someone else) accidentally twists the facts to their advantage.

Let's clear something up here. There's fact-checking, and there's error-checking. Fact-checking happens before and after the show, when it's easier to corroborate the data with the source. Error-checking is far more difficult, especially when you're rushing to get the graphic on air and have no time to do the fine proof-reading, especially if you are ignorant of the data. If someone is aware of the truth and fails to notice an error because of that, fine, that's confirmation bias, but your average control-room operator isn't always aware of what's correct and what's not correct.

This leaves open the possibility that the fact-checker - the producer writing the story - is either deliberately falsifying data or has a confirmation bias. I will admit that that's a greater possibility further up the line.

Dalton? Please, chill. This is not a slur on your profession, and I for one do understand that of the hands through which each individual bit of information on a news show passes, the vast majority are in no position to catch mistakes.

It doesn't affect my argument, which is a statistical one: if there is even one person who is in a position to spot errors, and that person is biased, or if there are multiple people in such a position and even one of them is biased... the rate of errors that favor the bias goes up. If the bias is strong or widespread, the rate of errors that favors the bias will skyrocket.

And that's before we even consider the possibility of any kind of intentional falsification, which I am leaving out of this whole question entirely.

OK?

Do I seem agitated? I'm just having a discussion.

You seem... a bit defensive. I'm not surprised, since your take on my position is based on something like "he's going on about how screwups in Fox prove that they are biased when he doesn't understand how easy it is to make screwups."

This is a misunderstanding. My argument is that given random screwups that happen all the time, that are sometimes but not always spotted and corrected before they air, we can infer things about the biases of the people who are in a position to spot and correct the errors. We infer this by looking at the patterns of errors, and seeing if some kinds of errors are made more often than others.

For example, if spelling errors are made a lot more often than normal on one channel, maybe that channel has a few people responsible for writing the bottom third who can't spell, and thus don't spot spelling errors. They generate more errors, and detect fewer of them, and therefore there are more errors to slip past the rest of the team. Even if the rest of the team (somehow!) still manages to spot and fix 95% of the original errors, if you double the number of original errors, you double the number of errors that get past the team.

Any individual spelling error is a random event that proves nothing. A pattern of consistent spelling errors tells you something- namely, that spelling errors are not being detected and corrected: someone is making more mistakes than normal, or spotting and fixing fewer mistakes than normal, or both.

The same goes for political bias, for the same reasons.

The main reason why I don't buy your confirmation bias theory is because a person has to be in the position of potentially knowing there's an error in the first place. This simply isn't the case. I will admit that in some cases errors might slip through with the person full-well knowing, if they noticed, that it was wrong - but in many cases nobody will realize, even for a long time, that there was a mistake.

Deliberate falsification is another story. There are dozens of unscrupulous people on both sides that I am sure feed biased data down the line. However, there is always room for doubt.

Again, my entire model of how this works treats the process of creating the segment as a black box. It does not matter how many people are involved; we can automatically ignore people who are unaware of an error, or unable to correct it. When I talk about "Larry, Moe, and Curly" as potential error-spotters, I am leaving out a ton of other people (say, Groucho, Harpo, Chico, and Zeppo) who cannot spot the error- either because they don't know the facts or because they don't have time to fix anything, or both.

If Harpo doesn't spot the mistake because he didn't have time, it is not his responsibility to spot the mistake, and he should not be treated as responsible. But Harpo's not being involved in error-checking is a constant: no matter what error is made, no matter who it favors, Harpo's not going to spot it. Therefore, Harpo can be abstracted out of the process of error detection, since he has no effect on it one way or the other.

What matters are the specific individuals who are responsible for making sure the news material is free of errors, who do have time to spot mistakes before they air, who do have access to the relevant facts. In this case, Larry, Moe, and Curly.

If one or more of those three is biased, that bias will have a dramatic effect on error detection rates, and you will see an imbalance in the number of errors made- because errors that exploit Curly's political blind spot are more likely to make the news than errors that don't.

[Alphawolf55]

If Fox News doesn't rely on lying, then why has Harper failed attempts to change the law to make lying on news channels legal kept Fox out of Canada?

[Dalton]

Simon, just to boil this down a bit since it's becoming somewhat tl;dr, you're saying that if someone who's error-checking knows definitively what the facts are, a confirmation bias is going to increase the chances that they'd have a "blind spot" in checking errors if those errors happen to support their viewpoint, yes?

Regarding the defensiveness, it may just seem that way since I'm one of those error-checkers.

[Simon_Jester]

Simon, just to boil this down a bit since it's becoming somewhat tl;dr, you're saying that if someone who's error-checking knows definitively what the facts are, a confirmation bias is going to increase the chances that they'd have a "blind spot" in checking errors if those errors happen to support their viewpoint, yes?

...More or less. If someone who's error-checking knows the facts, confirmation bias can make it more likely they simply won't see it: will not notice that someone has made a mistake because important bits of their brain do not perceive ti as a mistake- though in that case, if it's pointed out to them later, they'll go "shit, how did I miss that?"

Moreover, if someone sorta knows the facts, but doesn't have them in the forefront of their brain, confirmation bias makes it less likely they'll remember the facts correctly. If they don't know all the facts, and have to use their judgment when it comes to deciding to take the time to look things up, confirmation bias makes it more likely they'll go "Meh. Looks right" and give it a thumbs-up.

So when all is said and done, bias will make any given person who does error-checking in a news team less effective at error-checking. That would allow more errors to get through to the broadcast, and those extra errors would tend to favor whoever the person is biased in favor of. It can do this in many ways, even without any deliberate, systematic attempts to falsify news content.

If this happens a lot in a given news organization, the lion's share of the blame should go to whoever is deliberately encouraging bias among the news teams. In the case of Fox, that's going to be the people who make hiring decisions and have real influence over the corporate culture- they are the ones who allow their newsroom to become contaminated with wingnuts and true-believers who will predictably miss this kind of stuff, even if they're not consciously trying to skew their news coverage.

[FaxModem1]

Could we have the Fox News debate in a separate thread?

[Serafina]

Simon, just to boil this down a bit since it's becoming somewhat tl;dr, you're saying that if someone who's error-checking knows definitively what the facts are, a confirmation bias is going to increase the chances that they'd have a "blind spot" in checking errors if those errors happen to support their viewpoint, yes?

Regarding the defensiveness, it may just seem that way since I'm one of those error-checkers.

I would just like to point out something here: Someone who is checking for errors NEEDS to know the facts.

Now an interesting question is: What sorts of facts is he provided with?
One explanation could simply be that Fox News is providing inadequate facts for it's error-checkers. Maybe they have to look for inconsistencies on their own in some cases - in which case confirmation bias would play a large role - "hmm, though so, everyone is against commie-unions, looks right to me".
Now i don't have a lot of experience about how such things work, so that's just speculation. Still, Fox DOES err a lot and very often in favor of democrats - if the errors were random, then it should be 50:50.

[jcow79]

You claim that Fox admitted to "lie and distort the truth" and that it's "their right to do so". That's a wild exaggeration of the above and looks more like your specific takeaway from the incident.

I will admit as a definitive claim it came off as hyperbolic. However it certainly isn't MY specific takeaway. It was cited earlier in this very thread (p6) that their legal defense had "argued that they have the right to lie and call it news" which in my opinion is a tacit admission of lying. My claim was made in good faith as I felt it had already passed muster in the thread. If you feel the assertion is an exaggeration I gladly retract the claim as it's unnecessary to my point. There is plenty of other evidence of Fox being misleading or outright lying without requiring their confession.

That makes you automatically biased in your evaluation, since you are no longer willing to consider other factors simply because of the organization that made the error. There is always doubt.

Actually, I said I was NOT UNWILLING to consider other considerations that could cause such an error but unwilling to give them the benefit of the doubt. For me, their pattern of dishonesty shifts a greater burden of proof to them. Why wouldn't you consider the organization that made the error when making such considerations? If organization X has a reputation of dishonesty and organization Y has a reputation of integrity, would you consider them equal? I would want to factor in additional information when drawing my conclusions.

Here's the crux of my argument. Your initial claims were that this was likely a minor graphic error which I disputed, citing the fact the error also made it into the script. To this you expanded upon:

It's from the same source. That segment was written by a producer, who ordered a graphic to go with the story. If the producer had bad data, it would be disseminated to both the host via the script (who more or less goes by the prompter) and the designer (who does what they're told).

and to Simon you said

This leaves open the possibility that the fact-checker - the producer writing the story - is either deliberately falsifying data or has a confirmation bias. I will admit that that's a greater possibility further up the line.

Simon made the argument far better than I but it sounds like we agree. The difference being it seems you are extending them some professional courtesy about the cause of the error where I feel their pattern of dishonesty does not require my giving them the benefit of the doubt. As you said, doubt does always exist, but in this matter, I don't feel I owe them the benefit of it.

Off topic: I would like to add that I also feel the Fox news portion of this thread may deserve a split so the importance of the Wisconsin discussion is not lost.

[Edi]

This thread has been split from here.

[Dalton]

Apologies for the delay in returning to this discussion; I have been very busy.

Simon, just to boil this down a bit since it's becoming somewhat tl;dr, you're saying that if someone who's error-checking knows definitively what the facts are, a confirmation bias is going to increase the chances that they'd have a "blind spot" in checking errors if those errors happen to support their viewpoint, yes?

...More or less. If someone who's error-checking knows the facts, confirmation bias can make it more likely they simply won't see it: will not notice that someone has made a mistake because important bits of their brain do not perceive ti as a mistake- though in that case, if it's pointed out to them later, they'll go "shit, how did I miss that?"

Moreover, if someone sorta knows the facts, but doesn't have them in the forefront of their brain, confirmation bias makes it less likely they'll remember the facts correctly. If they don't know all the facts, and have to use their judgment when it comes to deciding to take the time to look things up, confirmation bias makes it more likely they'll go "Meh. Looks right" and give it a thumbs-up.

So when all is said and done, bias will make any given person who does error-checking in a news team less effective at error-checking. That would allow more errors to get through to the broadcast, and those extra errors would tend to favor whoever the person is biased in favor of. It can do this in many ways, even without any deliberate, systematic attempts to falsify news content.

If this happens a lot in a given news organization, the lion's share of the blame should go to whoever is deliberately encouraging bias among the news teams. In the case of Fox, that's going to be the people who make hiring decisions and have real influence over the corporate culture- they are the ones who allow their newsroom to become contaminated with wingnuts and true-believers who will predictably miss this kind of stuff, even if they're not consciously trying to skew their news coverage.

If that's indeed the case. As I think about it, day to day decisions on what stories to cover rely on few people, and those people are the ones most likely to have an agenda - theoretically speaking. In practice, they might just toe the party line to get and keep a job. As a personal anecdote (names have been removed to protect myself): Billy was a producer of a show at FNC known for being right-leaning, and now is a producer for a show on MSNBC known for being left-leaning. Alfred is a conservative Republican, but is the show director for an outspoken liberal. Et cetera.

I will admit as a definitive claim it came off as hyperbolic. However it certainly isn't MY specific takeaway. It was cited earlier in this very thread (p6) that their legal defense had "argued that they have the right to lie and call it news" which in my opinion is a tacit admission of lying. My claim was made in good faith as I felt it had already passed muster in the thread. If you feel the assertion is an exaggeration I gladly retract the claim as it's unnecessary to my point. There is plenty of other evidence of Fox being misleading or outright lying without requiring their confession.

I agree on that point - the Moody memos are evidence of that.

Actually, I said I was NOT UNWILLING to consider other considerations that could cause such an error but unwilling to give them the benefit of the doubt. For me, their pattern of dishonesty shifts a greater burden of proof to them. Why wouldn't you consider the organization that made the error when making such considerations? If organization X has a reputation of dishonesty and organization Y has a reputation of integrity, would you consider them equal? I would want to factor in additional information when drawing my conclusions.

Here's the crux of my argument. Your initial claims were that this was likely a minor graphic error which I disputed, citing the fact the error also made it into the script. To this you expanded upon:

It's from the same source. That segment was written by a producer, who ordered a graphic to go with the story. If the producer had bad data, it would be disseminated to both the host via the script (who more or less goes by the prompter) and the designer (who does what they're told).

and to Simon you said

This leaves open the possibility that the fact-checker - the producer writing the story - is either deliberately falsifying data or has a confirmation bias. I will admit that that's a greater possibility further up the line.

Simon made the argument far better than I but it sounds like we agree. The difference being it seems you are extending them some professional courtesy about the cause of the error where I feel their pattern of dishonesty does not require my giving them the benefit of the doubt. As you said, doubt does always exist, but in this matter, I don't feel I owe them the benefit of it.

That is definitely a consideration to take into account. Really, I'm playing devil's advocate here (and what a devil it is). I just wanted you to bear in mind that there are a lot of people both below the line and above the line who work their ass off just to get this shit on the air, and that a very few should bear the brunt of these errors - and also to remind people that sometimes a mistake is just a mistake, regardless of confirmation bias or errors missed.

[TheHammer]

I think while everyone understands that mistakes will be made in production of graphics, it seems like those mistakes almost always tend to be in favor of a right wing agenda. Is it possible that we only hear or see the graphics that look favor the right and there are in fact other graphic slip ups favoring the left? I suppose so. Are the number of graphic errors on Fox News roughly consistent with number of errors on other news networks? I don't have an answer to that, and I don't know that anyone has concrete facts in that regard.

[Rahvin]

Simon, this theory of yours is interesting, but I'd like to see some studies that political bias would make it less likely for someone to notice errors that agree with their point of view.

I'm sorry Dalton, but you're familiar with the concept of confirmation bias, aren't you?

When we see evidence that agrees with our existing model of the world, we don't tend to check it very thoroughly if at all. It's not unusual, and garners no special attention.

For research - check out Peter Wason's work from the 60's on the subject. Once a person attaches to a hypothesis, they largely look for confirmation of that hypothesis rather than falsification.

That's just one of the more famous experiments.

“If one were to attempt to identify a single problematic aspect of human reasoning that deserves attention above all others, the confirmation bias would have to be among the candidates for consideration. Many have written about this bias, and it appears to be sufficiently strong and pervasive that one is led to wonder whether the bias, by itself, might account for a significant fraction of the disputes, altercations, and misunderstandings that occur among individuals, groups, and nations.”

- Raymond S. Nickerson, psychologist and founder of The Journal of Experimental Psychology: Applied

A brief recap of several other experiments from Youarenotsosmart:

In a 1979 University of Minnesota study by Mark Snyder and Nancy Cantor, people read about a week in the life of an imaginary woman named Jane. Throughout the week, Jane did things which showcased she could be extraverted in some situations and introverted in others.

A few days passed. The subjects were asked to return.

Researchers divided the people into groups and asked them to help decide if Jane would be suited for a particular job. One group was asked if she would be a good librarian; the other group was asked if she would be a good real-estate agent.

In the librarian group, people remembered her as an introvert. In the real-estate group, they remembered her being an extravert. After this, when they were asked if she would be good at the other profession people stuck with their original assessment, saying she wasn’t suited for the other job.

The study suggests even in your memories you fall prey to confirmation bias, recalling those things which support your beliefs, forgetting those things which debunk them.

An Ohio State study in 2009 showed people spend 36 percent more time reading an essay if that essay aligns with their opinions.

Another study at Ohio State in 2009 showed subjects clips of the parody show “The Colbert Report,” and people who considered themselves politically conservative consistently reported “Colbert only pretends to be joking and genuinely meant what he said.”

Confirmation bias is very well-established, and it's essentially par for the course in human thought unless conscious effort is expended to overcome it.

It's not just possible, it's likely that, if[/i] we observe a pattern of error that swings overwhelmingly in support of one faction over the other, we are observing an effect of confirmation bias, where errors in favor of a preferred position are overlooked, while errors that would put a negative light on the favored position are corrected.

That, or it's a conspiracy, and I think you've very well debunked that.

ABE - The obvious counterpoint is that we are also affected by confirmation bias, and may only be noticing or recalling those instances of Republican-favoring errors on behalf of Fox News, ratehr than establishing an accurate model of the real pattern of errors. This is why anecdotes are bad, of course. I have no idea where we could get a statistical view of the errors made by Fox News and tally up those in favor of Republicans vs. those in favor of Democrats. Considering their hard-right-leaning commentators and utter lack of real representation from any other perspective, however, let's just say that I wouldn;t be at all surprised to find exactly the pattern we all think is there anecdotally.

[Dalton]

Simon, this theory of yours is interesting, but I'd like to see some studies that political bias would make it less likely for someone to notice errors that agree with their point of view.

I'm sorry Dalton, but you're familiar with the concept of confirmation bias, aren't you?

When we see evidence that agrees with our existing model of the world, we don't tend to check it very thoroughly if at all. It's not unusual, and garners no special attention.

For research - check out Peter Wason's work from the 60's on the subject. Once a person attaches to a hypothesis, they largely look for confirmation of that hypothesis rather than falsification.

Admittedly, I am not well-versed in the concept, though I do know that people tend to see what they want to see. But I do thank you for your examples.

Confirmation bias is very well-established, and it's essentially par for the course in human thought unless conscious effort is expended to overcome it.

It's not just possible, it's likely that, if[/i] we observe a pattern of error that swings overwhelmingly in support of one faction over the other, we are observing an effect of confirmation bias, where errors in favor of a preferred position are overlooked, while errors that would put a negative light on the favored position are corrected.

That, or it's a conspiracy, and I think you've very well debunked that.

ABE - The obvious counterpoint is that we are also affected by confirmation bias, and may only be noticing or recalling those instances of Republican-favoring errors on behalf of Fox News, ratehr than establishing an accurate model of the real pattern of errors. This is why anecdotes are bad, of course. I have no idea where we could get a statistical view of the errors made by Fox News and tally up those in favor of Republicans vs. those in favor of Democrats. Considering their hard-right-leaning commentators and utter lack of real representation from any other perspective, however, let's just say that I wouldn;t be at all surprised to find exactly the pattern we all think is there anecdotally.

An excellent point. We know of a pattern of FNC errors that are in favor of one side, but is there a pattern of the opposite? Or does a combination of both confirmation bias and internet overexposure skew those results?

[Korvan]

An excellent point. We know of a pattern of FNC errors that are in favor of one side, but is there a pattern of the opposite? Or does a combination of both confirmation bias and internet overexposure skew those results?

You don't even need to look for a pattern of the opposite, you just need to compile an overall error rate for R/D mismatches and then compare that to the error rate for republican scandal stories. If the rates are similar, then you have a single process producing the errors, but if you find the error rates do not match up, then you can conclude that two different processes are involved to produce the banners, one for general stories and one for republican scandals. This will not absolutely prove that the errors are being introduced deliberately, but I can't think of any other processes that would realistically account for the difference.

Now, compiling the data is no easy feat as you would need to source the banners as they were first run (to account for the errors being corrected on subsequent runs), and I imagine there is quite a few to go through.

[Dalton]

It would be quite difficult. There are thousands of banners on any given day, and the fixed items overwrite the incorrect items.

[mingo]

the only problem is when the cowinkidinks just pile up like cars in a Robert Zemekis film, it starts looking suspicious.

It's still circumstantial. You can't prove that all incidents of errors are part of an agenda.

Dalton, why are you working so hard at defending FNC? It seems they make more "mistakes" than anybody else and those mistakes never make Republicans look bad.

[Dalton]

the only problem is when the cowinkidinks just pile up like cars in a Robert Zemekis film, it starts looking suspicious.

It's still circumstantial. You can't prove that all incidents of errors are part of an agenda.

Dalton, why are you working so hard at defending FNC? It seems they make more "mistakes" than anybody else and those mistakes never make Republicans look bad.

"Seems". Did you actually read any part of this thread?

[mingo]

All right, Name one instance where a FNC "mistake" was in anyway bad for a Republican. And answer the question, why defend them? Splitting hairs on semantics doesn't make obvious lying not lying.

[Alyeska]

All right, Name one instance where a FNC "mistake" was in anyway bad for a Republican. And answer the question, why defend them? Splitting hairs on semantics doesn't make obvious lying not lying.

Dalton already pointed out that the same claim of Confirmation Bias could affect people who want to oppose FNC. The mistakes that paint Fox in a bad light are the ones mostly spotted.

You assume Fox only makes mistakes that benefit Fox. Which means you assume Fox makes no mistakes at all and is deliberately nickle and diming with small meaningless graphical pieces.

Did you even read the F*cking thread? Dalton WORKS in the news business and is arguing from a position of experience. Dalton works for MSNBC. And you accuse him of defending FNC?

[Dalton]

All right, Name one instance where a FNC "mistake" was in anyway bad for a Republican. And answer the question, why defend them? Splitting hairs on semantics doesn't make obvious lying not lying.

Onus of proof is on you. You claimed that Fox makes a lot of mistakes, all of which benefit Republicans. Back it up.

[Rahvin]

All right, Name one instance where a FNC "mistake" was in anyway bad for a Republican. And answer the question, why defend them? Splitting hairs on semantics doesn't make obvious lying not lying.

Can you honestly say that you've been taking an accurate record of any and all mistakes made by Fox? I certainly don't have a real tally from which to base a statistical analysis - all I have is anecdotal evidence, those cases where I have observed or been pointed to cases where Fox has mislabeled Republicans as Democrats when in trouble. I don't generally watch Fox myself, and I certainly don't watch it with a tally sheet waiting to catch errors.

As to the question of defense, that should be self-evident. Do you want an accurate view of Fox? Or do you want to believe that they always make errors in favor of Republicans, regardless of whether it's true or not? Anecdotally it sure seems to be pretty true, but the only way to find out if we're right is to test the hypothesis and obtain real statistical data. Personally, if Fox is as biased as we all think, then I want to believe that Fox is biased; conversely, if Fox is not as biased as our anecdotes suggest, then I want to be corrected. It seems to me that basing one's beliefs on accurate data rather than popular opinion is simply a matter of rational sense, and that suggesting we shouldn't at a minimum play Devil's Advocate and check our fucking facts is a good way to start down the road of becoming a delusional idiot.

Which is more important: believing that which is comfortable and pleasant, or believing what is accurate? When you flinch away from a suggestion that one of your beliefs may be wrong, when you specifically avoid attempts to test the belief, you have a problem. This question isn't one of semantics, its about finding out what's actually happening, whether reality matches what all of our anecdotes suggest.

If we were to actually watch Fox for a year, fact-check everything, and tally the errors and which if any political party benefited from each, and we discovered that Fox actually makes errors in favor of Democrats roughly as frequently as in favor of Republicans and we just never noticed, what then would you say? I'd be rather surprised myself, the anecdotes are just too numerous to regard a hypothesis of bias as unlikely, but I'd shift my beliefs in an instant if we discovered that the anecdotes were just wrong. Don't we make fun of Creationists and other retards for bringing up their anecdotal evidence of the supposed power of prayer in the face of double-blind experimental results clearly and consistently showing no statistically significant effect?