Bias in news sources (Split from: Impeachment Hearings)

[zapatos]

10 hours ago, J.C.MacSwell said:

Hyperbole, but more than a grain of truth to it with regard to much of current media.

 

YJ02 is a big boy and has had every opportunity to clarify his assertion. I don't think he needs you speaking for him.

And if it was simply hyperbole and he only now chooses to tell us then we are again wasting our time. What is the point of having a debate if you can later say "I know I let you think that while we were talking, but only now will I tell you what I really meant."

[CharonY]

18 hours ago, YJ02 said:

Let me give you another example if I could.

These linked to polls show that their has been a large increase among african americans , and non whites, for Trump over the last 6 months or so.

As much as a 30% approval rating for Trump is indicated (to which i call BS btw)

Now keeping in mind your statement of - "If the polls are biased, you must show where, why, amd how. You can’t summarily dismiss them because of the source. "

so, do you, or anyone here, accept them or dismiss them?

extract from: https://pjmedia.com/election/and-then-there-were-3-marist-poll-shows-33-approval-for-trump-among-non-white-voters/

"A new NPR/PBS NewsHour/Marist Poll concludes that 33 percent of non-white voters approve of Donald J. Trump's performance as president. 57 percent of non-white voters supposedly disapprove of his performance. Although these would be horrendous approval ratings for a Democrat, they're actually very good for a Republican. For example, back in 2005, African American approval of then President George W. Bush was a mere 16 percent.

As we reported on Monday, both a Rasmussen poll and a poll by Emerson show 34 percent approval for President Trump among African Americans."

There are two sources of information here and they are not equal, though they are treated as one unit. The first are the polls themselves. The article does not link to the source (only to the one regarding Bush, but more about that in a bit). So the first question to ask is whether the mentioned numbers are accurate. The article itself is clearly not a news article (calling someone from CNN a crybaby, and making highly selective extrapolation based on the data). 

In other words, the issue with that snipped would not only be whether the numbers are accurate but also whether the extrapolations are. Let's discuss the data first. The challenge with polls or any other data is first the methodology. The issue of course is that a certain level of expertise is required. There are different measures that could be used, but Marist generally is considered a very good polling organization. Here, the article is showing the bias part by declaring that 33% non-white support must be a high African American approval rating, but the poll does not provide that information. Going back it shows that this particular poll has shown roughly that level of approval non-whites. So there actually has not been a lot of change in the numbers. Going deeper into sample size it shows that among the ~1k persons polled there were roughly 100 African American (a bit more Latinos and others). So specifically tracking the approval from that small group is going to be difficult, which is probably why they only present it in aggregate. This limit in is also reflected by the fact that in the poll the white group is dissected in more detail (e.g. education gender etc.) whereas they did not perform any further distinction in the non-white set. In other words, the data set is good to track overall approval, but probably not a good set to specifically trace African American sentiments. So here, one could ask whether a poll with a larger African American respondent base would show the same results (and a quick look across other polls does not seem to make it likely).

Rasmussen is the one showing quite high African American support, but in order to figure out whether either makes sense one would have to look at other available information. For starters, ~8% of African Americans have voted for Trump in 2016 and most polls directly asking African Americans are at a consistent approval rating between 8-12% (more or less) Rasmussen explain that their methodology is different is focused on likely voters. Yet that would imply an more than 20% jump that other methodologies did not capture. So it could be that this method is great (if the outcome turns out to be true) or that it is, in fact faulty (I cannot quickly find details regarding their sample composition, for example). But to re-iterate a previous point, it is well-known fact that general polls with usually 1-2k responses tend to undersample (statistically) minorities if they use the actual composition of the US-population, resulting in imprecise data. In order to figure out how the distribution truly are, you generally have to oversample and then perform a weighing correction.

It does not mean that one needs to reject data outright, just that one needs to be careful in the interpretation and contextualization. I.e. the methodology determines what the data actually might be able to tell us. Rather obviously the linked article has not such qualms and provides a strong narrative without proper contextualization

[iNow]

1 hour ago, CharonY said:

calling someone from CNN a crybaby,

Technically, she's a former GOP strategist who worked quite a bit with Jeb Bush who sometimes presents now on CNN during various panels, but I digress... Good summary.

[CharonY]

I think this illustrates also the issue with the democratization of information and social media. Properly interpreting data is time consuming and requires expertise. However, falsehoods and narratives can be fabricated almost instantaneously. As such, the infosphere almost by default will contain fare more false claims relative to true ones. Putting in the argument that every source is biased or false further dilutes the value of actually well-analyzed information. After all, why put the effort in, if it ends up against weighed against tons of crap?

There is an effort to teach kids about source criticism. But the issue in the end is that there has to be a certain level of trust. And that trust has been systematically eroded which ultimately (in my mind at least) puts democracy at risk.

[iNow]

14 minutes ago, CharonY said:

the issue in the end is that there has to be a certain level of trust. And that trust has been systematically eroded which ultimately (in my mind at least) puts democracy at risk

One side is arguing about how each individual chess piece is allowed to move while the other side is drilling giant holes from underneath the board until it’s nothing but pulp. It’s a sort of culture-wide Gish Gallop. 

[YJ02]

19 hours ago, J.C.MacSwell said:

 INow. It is not all or none.

Where do you get that YJO2 claims they are useless and biased by default?

They are to some degree suspect.

This is to some extent true of climate change data (an example you brought up that no one is describing as a red herring...)

It was an example. No one claimed you were obligated to respond to it. The facepalm was you not allowing the same courtesy.

 

No we can't.

So, first, sorry I am not around on my PC all day just waiting to fire back a response. I know I am retired, but there are still things to do

Polls- the data is not biased, its data. How can numbers be biased? Numbers are useful.

it is the interpretation of the numbers, the narratives that are written as a result of them. Then the spin-or not- that a right leaning or left leaning pollster may put on the meaning of the numbers. They may over inflate the importance or they may try to undercut the importance of the results.

The article I attached, referenced CNN's coverage of the poll results. Navarro's response- invoking Kanye West and some other well know afr-am Trump supporters as being the sample population-- is a example of bias in newscast. If CNN was truly concerned with fairness and objectivity, then her response should have been edited on delay or however its done. Allowing her to include her opinion, heavily tainted by personal biases she holds, should be unacceptable 

--------------------------------------------------

Polls are not entirely useless, they are very useful to those seeking out confirmation of their own beliefs. If a person is a Sander's supporter for example, what polls do you think they will first search out and then read? Most likely ones that are titled something like "What is your opinion of Bernie Sanders?" 

When, if they wanted a more accurate report of the standing of Sanders, they should be looking for a poll that includes all democratic candidates.

So, personal bias can ,and does, affect the news/polls we seek out in the first place, thereby opening us up to what amounts to seeking out validation.

So, yes they are useful in that regard. 

Too many people will not look for views or opinions from the other side. Many will actually take active measures to avoid seeing anything that might be against their views. It is the way things are going more and more today.

It is strongly connected to the "shout down/shut down" tactic used by people on both sides of the argument these days. They won't even let a person speak or read their full post or article simply because they may have a more moderate view or they may say one or two positive things about what/who you dislike. It is dangerous info to them;  they don't want to be challenged by other views/opinions/data. Better to shut it down before it is even said/heard/read/viewed

Assuming one is a "MAGA hat wearing Trump supporter" just because they question how the impeachment is being done (NOT because it is being done) is a further symptom of this social phenomena which is actually a process involving several layers of bias.

[zapatos]

3 minutes ago, YJ02 said:

Assuming one is a "MAGA hat wearing Trump supporter" just because they question how the impeachment is being done (NOT because it is being done) is a further symptom of this social phenomena which is actually a process involving several layers of bias.

That would be a terrible thing. Does it actually happen?

[YJ02]

As for the news. I grew up watching stuff like Walter Cronkite and a young Dan Rather.

In those 30 min broadcasts, the actual news was delivered with out comment or opinion.No 'slant' or bias made its way in to the reporting. But the anchors did have a chance to speak. At the end of every broadcast we all knew the 'editorial' segment was coming.If we didn't want to hear opinion, we could then shut it off having already gotten the news of the day.

Now its an hour or more of the same show  where snippets of news are used as springboards for a barrage of biased opinion. Places like CNN and FOX used to have on a guest representing the opposing view. Now that is rare. I don't have "live tv" but in the segments of American news I do sample on Pluto or youtube, I do not see any 'point - counterpoint' type discourse. It is all one sided with guests all from the left or the right and the only differences between them being the degree in which they are left or right.

DW, France24, BBC, even Al Jazeera are what I normally select to watch--yes, I am biased against American "news" sources

[iNow]

12 minutes ago, YJ02 said:

Polls are not entirely useless, they are very useful to those seeking out confirmation of their own beliefs.

<snip>

So, personal bias can ,and does, affect the news/polls we seek out in the first place, thereby opening us up to what amounts to seeking out validation.

So, yes they are useful in that regard. 

Implicit in your comment here is the idea that this is the ONLY way they’re useful. Is that what you intended to convey?

11 minutes ago, zapatos said:

That would be a terrible thing. Does it actually happen?

In fairness, yes. It does in some places. Social media is often a cesspool on all sides. 

[YJ02]

4 hours ago, iNow said:

Implicit in your comment here is the idea that this is the ONLY way they’re useful. Is that what you intended to convey?

Yes, in today's socio-politico climate in America.

I would even suggest (not claim definitively) that the failure of polls in accurately predicting outcomes in the 2016 Presidential election, and in the 2012 election to a lesser extent, may have been due to several levels of bias. By the pollsters, the participants, and those charged with analysis.

I also think that in those instances-and today, the news organizations pick what polls to report on based on what message they want to convey--another symptom of the current 'news as entertainment' model.

 

5 hours ago, zapatos said:

That would be a terrible thing. Does it actually happen?

What? To question HOW the government is doing something?

I suppose to some, that would be a catastrophic occurrence.

[J.C.MacSwell]

MSNBC has no mention of Andrew Yang since Nov 26

http://www.msnbc.com/search/andrew yang

I wonder how long they can keep it up?

[iNow]

23 hours ago, J.C.MacSwell said:

I don't get the impression anyone is claiming polls are, by default, useless.

We can all read the same words. Why is it that you insist on mischaracterizing in this manner

Given my most recent exchange with the OP, your impression was clearly wrong and I mischaracterized nothing. Thanks for making it personal, though

6 hours ago, YJ02 said:

I would even suggest (not claim definitively) that the failure of polls in accurately predicting outcomes in the 2016 Presidential election, and in the 2012 election to a lesser extent, may have been due to several levels of bias. By the pollsters, the participants, and those charged with analysis.

I saw no polls which gave Trump zero chance of winning. Most put his chances around 30%, and while that’s a lower probability than a candidate would like it, it’s still a 3 in 10 chance of winning and shouldn’t be conflated with the poll being either wrong or biased in the way you’re suggesting. 

6 hours ago, YJ02 said:

What? To question HOW the government is doing something?

No. As should be clear with an honest reading of the exchange, he was asking about individuals on social media automatically calling those who question the current government a “MAGA hat wearing Trump supporter.” I even supported you on this point, but your response unfortunately suggests honest discourse may not be your intent here. Sad. 

Edited December 11, 2019 by iNow

[zapatos]

8 hours ago, YJ02 said:

What? To question HOW the government is doing something?

 

No, to assume one is a "MAGA hat wearing Trump supporter" just because they question how the impeachment is being done. It's getting old.

Edited December 11, 2019 by zapatos

[MigL]

I know I'm often guilty of the same thing, but if INow and Zap on one side, and YJ02 and JC on the other, would pause before re-posting and ask for clarification of a comment made by the other side, instead of assuming and bringing their own biases into play ( yes, we all have them ) this conversation/discussion would be a lot more congenial.

I think we all agree that D Trump is bad for the country ( and the rest of the world ), we just have a different perspective on which method/process for removal would do the least damage to the country and its government.

[zapatos]

21 minutes ago, MigL said:

I know I'm often guilty of the same thing, but if INow and Zap on one side, and YJ02 and JC on the other, would pause before re-posting and ask for clarification of a comment made by the other side, instead of assuming and bringing their own biases into play ( yes, we all have them ) this conversation/discussion would be a lot more congenial.

 

Of course. Always a good reminder.

[iNow]

57 minutes ago, MigL said:

I know I'm often guilty of the same thing, but if INow and Zap on one side, and YJ02 and JC on the other, would pause before re-posting and ask for clarification of a comment made by the other side, instead of assuming and bringing their own biases into play ( yes, we all have them ) this conversation/discussion would be a lot more congenial.

You mean exactly as I did 14 hours ago?

 

14 hours ago, iNow said:

Implicit in your comment here is the idea that this is the ONLY way they’re useful. Is that what you intended to convey?

 

[J.C.MacSwell]

25 minutes ago, iNow said:

You mean exactly as I did 14 hours ago?

 

 

Yes INow. That was much better.

Than this more recent one:

 

3 hours ago, iNow said:

 

No. As should be clear with an honest reading of the exchange, he was asking about individuals on social media automatically calling those who question the current government a “MAGA hat wearing Trump supporter.” I even supported you on this point, but your response unfortunately suggests honest discourse may not be your intent here. Sad. 

 

[iNow]

Your opinion is noted

[zapatos]

9 minutes ago, J.C.MacSwell said:

Yes INow. That was much better.

Than this more recent one:

Just had to take a potshot at him...

I guess congeniality won't be joining us today.

[YJ02]

On 12/11/2019 at 8:14 AM, iNow said:

I saw no polls which gave Trump zero chance of winning. Most put his chances around 30%, and while that’s a lower probability than a candidate would like it, it’s still a 3 in 10 chance of winning and shouldn’t be conflated with the poll being either wrong or biased in the way you’re suggesting. 

Ok.

I think what you may be inferring is that you are one of the few in America who actually looks at more then one poll, from both sides of the political spectrum?

But I hope we are honest enough to admit we know that the majority of Americans do not do this, I'd guess that they would not even know they could access polling data on their own.They just get the polling data from whatever news source they prefer (through whatever biased lens they report from).

Americans have no need to any longer have to read or view the news, consume the info, or consider any data in search of confirmation for their political bias- the polls and the media construct a ready made confirmation bias package for them. The "we'll do all the thinking for you, stay tuned in 24/7" news network. They only need to turn on the tv or go to the website-- apparently the self imposed limit of their free choice.

In this, then yes, polls are useful.

On 12/11/2019 at 9:57 AM, MigL said:

I think we all agree that D Trump is bad for the country ( and the rest of the world ), we just have a different perspective on which method/process for removal would do the least damage to the country and its government.

 So, I see that you must have a closet full of starched, bright red MAGA hats to wear-a new one for each day! 

 

[iNow]

7 minutes ago, YJ02 said:

I think what you may be inferring is that you are one of the few in America who actually looks at more then one poll, from both sides of the political spectrum?

No. I’m saying that none of the polls gave him a zero percent chance of winning hence your claim of bias in said polls is unfounded. 

9 minutes ago, YJ02 said:

Americans have no need to any longer have to read or view the news, consume the info, or consider any data in search of confirmation for their political bias- the polls and the media construct a ready made confirmation bias package for them

I suppose SOME news sources do this, but your implication is that ALL news sources do this, which is quite clearly a load of horseshit. 

[YJ02]

2 hours ago, iNow said:

I suppose SOME news sources do this, but your implication is that ALL news sources do this, which is quite clearly a load of horseshit.

Then please name me one American news source that is without bias and only reports news from a complete objective stance only.

I'll admit, and I said it someplace here, I do not have 'live tv', I don't get the 'big 3' networks. I used to and got very sick of the nonsense during the Obama admin. I now watch American news snippets/aggregates on places like Pluto Tv and Youtube as well as . I also listen to NPR and all other news comes via foreign sources as I listed.

I want to believe there is  Cronkite type news out there still, because America definitely needs one or a hundred.

thanks

[MigL]

No MAGA hats in my closet, YJ02.
I'm Canadian.

Besides hats mess up my hair.

[J.C.MacSwell]

2 minutes ago, MigL said:

No MAGA hats in my closet, YJ02.
I'm Canadian.

Besides hats mess up my hair.

If you're wearing all of them at once...no wonder!

[YJ02]

Found this tonight.... "binary Bias"

an excerpt from:https://www.psychologytoday.com/us/blog/psych-unseen/201910/can-polarized-american-politics-find-the-middle-way

Quote

 

Recent research by social psychologist Jan-Willem van Prooijen has determined that certain cognitive biases, like black-and-white thinking, are shared at both extremes of the political divide.2 The “cognitive simplicity” of black-and-white thinking, van Prooijen claims, often leads to overconfidence in our political beliefs.

What can we do to gain perspective beyond this inherent cognitive rigidity? The answer is hardly new. With its roots in early Buddhism, the philosophical prescription to seek “the Middle Way” has since been reflected in the complementary dualism of the I Ching’s “yin and yang” and in the more modern Western “thesis, antithesis, synthesis” approach of Hegelian dialectic

 

.more : https://journals.sagepub.com/doi/full/10.1177/0963721418817755

3 minutes ago, MigL said:

No MAGA hats in my closet, YJ02.
I'm Canadian.

Besides hats mess up my hair.

 there is always 'make Alberta great again'. you never know when some MAGA's may come in handy. and they already match the Canadian flag!

Winning all around! Tremendous assets!!