Hacker News Reader

Comment by scholia | Hacker News Reader

It's really not hard to identify partizan news sources, which is what this is based on. They often self-identify themselves.

It's really not hard for professional journalists to fact-check partizan articles, and I'd assume the ones hired by Buzzfeed used several well-known fact-checking sites, including Politifact.

> sounds like a way to abrogate the responsibility of the individual to understand and analyze sources

But of course it doesn't do that. It's just a way of getting a machine to identify partizan stories that are statistically likely to contain a higher proportion of lies, especially if they are from right-wing sources (Table 1).

As we know all too well, individuals commonly abrogate the responsibility to understand and analyze sources all by themselves.

View on HN

> But of course it doesn't do that. It's just a way of getting a machine to identify partizan stories that are statistically likely to contain a higher proportion of lies, especially if they are from right-wing sources (Table 1).

I think if a system like this got traction, then like many things academics produce, it will be misconstrued and misused.

> It's really not hard for professional journalists to fact-check partizan articles, and I'd assume the ones hired by Buzzfeed used several well-known fact-checking sites, including Politifact.

You should not assume things like that. And again, if they were just gonna look stuff up on Politifact, why couldn't the researcher do that themselves? It's clear there was a more in-depth process here. I think it is totally fair to say that the process could be a good one even if Buzzfeed is a bad choice of organization to run the process with, but further research would have to bear that out. Because only Buzzfeed was involved, it's impossible for us to determine whether there was a bias in the data because of who produced it.

That is, we can't be sure if that data would be similar if we had, say, the National Review doing fact-checking. That's why we need either (preferably) more than one fact-checking organization, or at least a single but highly neutral one.

> As we know all too well, individuals commonly abrogate the responsibility to understand and analyze sources all by themselves.

Which does not mean we shouldn't be careful about enabling it.

My fear here is the common trope of "X new discovery is an oracle," where we take some interesting tidbit about patterns (in in this case, in linguistics) and decide it is absolute truth or applies everywhere.

> You should not assume things like that.

On the contrary, this is a very reasonable assumption. You appear to be confusing opinions and facts.

You also appear to have an uninformed opinion of Buzzfeed News and the journalists it employs.

You also don't seem to understand the main point of the whole paper, which is that "the writing styles of otherwise opposing orientations, namely left-wing and right-wing, are in fact very similar: there appears to be a common style of extremism."

Based on my long experience as a newspaper journalist, this was already obvious to me and no doubt many other people. However, it won't make any difference to the significant proportion of the American public that doesn't read news for honest information, but only to prop up their existing and highly distorted views.

> On the contrary, this is a very reasonable assumption. You appear to be confusing opinions and facts.

It's not reasonable to imagine a study used methods just because you would have done so if it were you, even if you know that those methods are the best or most standard ones, then defend the study on the basis of the imaginary method. The point of having a peer-reviewed study is that it clearly indicates how the fact-checking was done.

> You also appear to have an uninformed opinion of Buzzfeed News and the journalists it employs.

On the contrary - I read them and am quite familiar with their work. You can look through my comment history and see me discussing them quite extensively just the other day. Generally speaking, their journalism work is excellent, but it does have somewhat partisan viewpoint in my opinion and they do not always act as an organization in ways that I support.

But again, even if Buzzfeed is not the worst choice, it certainly isn't the best one; you really aren't addressing the point I've brought up a couple times now that there are other organizations better suited to this particular kind of journalistic work.

> You also don't seem to understand the main point of the whole paper, which is that "the writing styles of otherwise opposing orientations, namely left-wing and right-wing, are in fact very similar: there appears to be a common style of extremism."

I understand the point. I just don't think it is useful, and even if it is, I don't particularly think the study has proven it beyond doubt.

> On the contrary - I read them and am quite familiar with their work. You can look through my comment history

Fair enough: I didn't look at your comment history...

> but it does have somewhat partisan viewpoint in my opinion

There is no view without a point of view, and all you're saying is that your own view is to he right of Buzzfeed's.

Meanwhile, individual journalists at Buzzfeed News also have views that are to the right and to the left of Buzzfeed's.

None of which has anything to do with the actual paper, which wasn't about views but about facts.

> There is no view without a point of view, and all you're saying is that your own view is to he right of Buzzfeed's.

Actually, my personal political beliefs are probably substantially to the left of Buzzfeed's. That doesn't make me incapable of identifying a partisan viewpoint. I can still tell, even if I mostly agree with it.

> None of which has anything to do with the actual paper, which wasn't about views but about facts.

It has everything to do with it. Partisan sources will rate other partisan sources on their side as more truthful than partisan sources on the other side. One possible explanation of table 1 (not necessarily the true one, but a possible one) is that the reason right-leaning sources were found to lie so often is because they said things the left-leaning fact-checkers didn't agree with, or that the left-leaning fact checkers were harder on the right-leaning sources than they were on the left.

Using Buzzfeed opens the study up to criticism and may detract from the value of the data as a result.

You're still missing the point. Facts are facts, whether you agree with them or not.

A statement of fact is only read as such when the reader and writer agree on the terminology used. That might lead to more interesting studies.

That's not actually true. Try reading some fact-checking sites.

Care to elaborate? How can people discuss facts when they have contradictory definitions of the words being used in an attempt to communicate?

In my experience, people don't actually have "contradictory definitions of the words being used". If they are mis-using a word, that's simply wrong.

But you're welcome to quote me an example from, say, Politifact or the Washington Post fact-checker or Snopes....

It's not about whether I agree with them. It's about whether, when someone has been charged with reviewing them, they agree with them. Such a person may be tempted to subvert or disallow them on the basis of their opinion. This burden is on this study to demonstrate that this has not happened.

> It's just a way of getting a machine to identify partizan stories that are statistically likely to contain a higher proportion of lies, especially if they are from right-wing sources (Table 1).

So, anything from the right should be a lie by default? Then google or whoever can just unrank the article to oblivion, since it statistically contains "lies" according to a biased majority? Do you see how dangerous this is?

No: see Table 1. It's possible to write partizan stories without using false facts. The Other 98% does it. However, the right-wing sources used -- Eagle Rising, Right Wing News, Freedom Daily -- had a higher proportion of false facts than the left-wing sources.

You must surely know that many right-wing sources are notorious for lying, especially Breitbart and Infowars. Being a right-wing source shouldn't mean your lies get a free pass.

For some interesting information on how right-wing sources are obliged to lie or be rejected by their target audience, see http://uk.businessinsider.com/conservative-media-trump-drudg...

I have NO PROBLEM with a social media site tagging a source as credible/un-credible. I can still choose to read or not read the story. I can look at the credibility rating, and wonder if A) the rating system is biased, and B) if a liar figured out how to game the automated rating system.

Bias can be dealt with by being an open system. People can see the algorithm, data input, and judge whether the output is fair. Independent audit and review of the system would be miles better than what we have now.

As far as bad-actors trying to game the system, that's another matter.

Having a rating system is just a tool. It will make life easier for every consumer of news.

Life with a spam filter is way better than no filter. We learned this in the 1990's after spending a decade writing them for email, when email became nearly useless as a tool because of spam.

Reply on news.ycombinator.com