Fake news sharing in US is a rightwing thing, says study

[smilesammich 10,727]

1 minute ago, Boiler said:

I know I know...

and maybe give ol joe a try 

[Boiler 45,481]

Love my joe in the morning.

[bcking 3,157]

For people who don't understand what research is. Here are some good books to pick up if you want to learn.

 

https://www.amazon.com/Educational-Research-Conducting-Quantitative-Qualitative/dp/0131367390

 

https://www.amazon.com/Essentials-Research-Methodology-Geoffrey-Marczyk/dp/0471470538

 

I'm not saying I like this particular study. Their methods could be clearer, and I would like more supplemental information. They also should have a paragraph on their "limitations", though they also don't really have a discussion at all. They just present conclusions.

 

The study is at significant risk of bias based on how they crafted their methodology. If I were reviewing it for publication I would want major edits. I wouldn't, however, write back to them saying "This isn't research, this is just an opinion piece.". I would lose any credibility I have as a reviewer. I wouldn't be asked to review papers for that journal again, most likely.

 

Even the studies a journal may reject for publication aren't necessarily "not research". I would potentially not support publishing in a journal I review for because of the risk of bias. Though honestly if they just added a couple of sentences it would probably be fine. One sentence in the methods for how they resolved differences between coders (majority rule, primary coder as a tie-breaking vote etc...). Then One sentence (or maybe 2-3) on limitations at the end. One major limitation being the validity of their tool for defining websites as "fake" since I don't recall them having a reference for the tool being used in the past. The fact that they report an inter-coder reliability for the 3 primary coders makes me hopeful that they could likely provide more information and make their methods more clear if they were asked to do so.

[The Nature Boy 7,659]

3 hours ago, smilesammich said:

don't be jealous of my superhero powers and please don't bring totally unrelated nonsense from other threads into this one. if anyone has a problem with what i said they can report it. k thx.

Reported draconian measures recommended 😂

[smilesammich 10,727]

8 minutes ago, Nature Boy Flair said:

Reported draconian measures recommended 😂

bon voyage - you'll miss me when i'm gone!

 

[Dashinka 24,492]

4 hours ago, bcking said:

Not true.

 

A study may have subjective results but was done in a systematic way, with methodology that was defined a priori and results that were testing a specified hypothesis. That is VERY different from people forming an "opinion" on the subject.

Doing a study in a systematic way does not make it objective.  The parameters they used to define a fake news site was purely subjective and as such subject to the bias of the folks doing the systematic study.

[bcking 3,157]

8 minutes ago, Bill & Katya said:

Doing a study in a systematic way does not make it objective.  The parameters they used to define a fake news site was purely subjective and as such subject to the bias of the folks doing the systematic study.

A research study doesn't have to be entirely objective though. Plenty of research is subjective.

 

My thesis involved identifying phenotypic changes in the gut cells of C. Elegans using microscopy. Very subjective. Also scientific.

 

They report inter-rarer reliability, so while subjective they did show that among multiple raters the choices were consistent.

 

Another example that comes to mind is the countless studies I read with echocardiogram interpretations (ultrasound of the heart). I actually reviewed a paper two weeks ago for a journal with that as part of their methods. Echo readings are subjective and subject to the reader making estimates and measurements.

 

Typically with echocardiogram results the first studies that use them publish inter-rater (and ideally intra-rater) reliability estimates.

 

While something may be subjective - three independent reviewers coming to the same conclusion definitely supports its use. Still subjective, but they are doing their due diligence.

Edited February 6, 2018 by bcking

[Dashinka 24,492]

9 minutes ago, bcking said:

A research study doesn't have to be entirely objective though. Plenty of research is subjective.

 

My thesis involved identifying phenotypic changes in the gut cells of C. Elegans using microscopy. Very subjective. Also scientific.

 

They report inter-rarer reliability, so while subjective they did show that among multiple raters the choices were consistent.

 

Another example that comes to mind is the countless studies I read with echocardiogram interpretations (ultrasound of the heart). I actually reviewed a paper two weeks ago for a journal with that as part of their methods. Echo readings are subjective and subject to the reader making estimates and measurements.

 

Typically with echocardiogram results the first studies that use them publish inter-rater (and ideally intra-rater) reliability estimates.

 

While something may be subjective - three independent reviewers coming to the same conclusion definitely supports its use. Still subjective, but they are doing their due diligence.

How do they make sure that the independent reviewers do not all share the same biases when assigning a rating for the subjective parameters?

[bcking 3,157]

2 minutes ago, Bill & Katya said:

How do they make sure that the independent reviewers do not all share the same biases when assigning a rating for the subjective parameters?

They don't specify.

 

Again I'm not saying it's a great study. As I already said I would not accept it if I was a reviewer for a journal, without revisions.

 

My issue was with someone else here who claimed that it wasn't a research study. It ABSOLUTELY is.

 

Your concern is incredibly valid. They could have explained more in their methods. They could have potentially partially blinded the review process to help minimise bias. At the end of the day though the bias of the reviewers in their assignment is a perfectly valid critique. 

 

It should have been listed in a limitations section. It's a critique, but it doesn't invalidate it as a research study.

[Dashinka 24,492]

1 minute ago, bcking said:

They don't specify.

 

Again I'm not saying it's a great study. As I already said I would not accept it if I was a reviewer for a journal, without revisions.

 

My issue was with someone else here who claimed that it wasn't a research study. It ABSOLUTELY is.

 

Your concern is incredibly valid. They could have explained more in their methods. They could have potentially partially blinded the review process to help minimise bias. At the end of the day though the bias of the reviewers in their assignment is a perfectly valid critique. 

 

It should have been listed in a limitations section. It's a critique, but it doesn't invalidate it as a research study.

Apparently though, it was accepted without revisions, or a more in depth explanation of their methodology.  To me, this makes the conclusions suspect.

[bcking 3,157]

Just now, Bill & Katya said:

Apparently though, it was accepted without revisions, or a more in depth explanation of their methodology.  To me, this makes the conclusions suspect.

I actually am not sure it was published in a peer reviewed journal. The link I found I believe was from the research group's website. I can check again, I'm just thinking from memory.

 

A proper analysis of the paper would say that it is at high risk of bias, so any conclusions are of low grade. As you point out an entirely different group of researchers could end up with very different results.

 

Still a research study though. That was my major point. Question the conclusions sure, especially since they don't list any limitations. But it is certainly not just an "opinion piece".

[Steeleballz 7,550]

 

    It's not the first study/research/opinion on this subject. I've seen 3 different studies posted as topics on this forum alone that said the same thing. I posted one of them. Over time, the consensus logically becomes less suspect when every study comes to the same conclusion. 

 

   I have not looked at this study in as much detail as the one I posted earlier, but I suspect a similar theme. They are not talking about CNN or FOX news. They are talking about the propensity of right wing news to spread on sites like Facebook or your uncles blog. If we don't like the political bias of CNN or FOX, that's one thing. The unsourced, unverified rubbish that shows up on some of these other sites it a whole different level. For whatever reason, the right has seen a disproportionate amount of these sites. Rather than get defensive about it, it would probably be better to ask why.

 

   

 

   

[eieio 1,594]

I'll make it easier for some. CNN is fake news.  

 

The study in this thread is basically trash.

[bcking 3,157]

37 minutes ago, eieio said:

I'll make it easier for some. CNN is fake news.  

 

The study in this thread is basically trash.

No offense, but that is spoken like somehow who likely knows very little about how to read scientific literature.

 

 

Edited February 7, 2018 by bcking

[eieio 1,594]

10 minutes ago, bcking said:

No offense, but that is spoken like somehow who likely knows very little about how to read scientific literature.

 

 

None taken. You watch CNN. Thats about all one needs to know.