Exposure to Ideologically Diverse Information on Facebook(research.facebook.com)
research.facebook.com
Exposure to Ideologically Diverse Information on Facebook
https://research.facebook.com/publications/1629735407303857/exposure-to-ideologically-diverse-information-on-facebook/
7 comments
> The most likely reaction to reading bullshit you disagree with is "This Is Bullshit"
I wonder what do people do when they read bullshit that they agree with.
Is Facebook trying to find ways to maximize how much bullshit will you agree with?
I wonder what do people do when they read bullshit that they agree with.
Is Facebook trying to find ways to maximize how much bullshit will you agree with?
> I wonder what do people do when they read bullshit that they agree with.
They agree with it (maybe while realizing it's bullshit).
I think among people calling for content diversity, there's a mostly explicit assumption that encountering more opinions you disagree with will cause you to either treat those opinions with more respect, or else smell bullshit on your own side of the fence more often.
IME it doesn't work that way, regardless of what happens in the lab. If anything, reading bullshit you disagree with has the opposite effect -- "wow, that's a really shitty reason to believe X. People who believe X must be stupid to believe that reason."
Seeing high quality justifications for opinions you disagree with or meeting people you respect who disagree with you can both lead to attitude changes and even changes of opinion.
But reading punditry on the Internet? Nope.
> Is Facebook trying to find ways to maximize how much bullshit will you agree with?
Well definitely, because most marketing is bullshit. But I don't think that's what this paper is about.
They agree with it (maybe while realizing it's bullshit).
I think among people calling for content diversity, there's a mostly explicit assumption that encountering more opinions you disagree with will cause you to either treat those opinions with more respect, or else smell bullshit on your own side of the fence more often.
IME it doesn't work that way, regardless of what happens in the lab. If anything, reading bullshit you disagree with has the opposite effect -- "wow, that's a really shitty reason to believe X. People who believe X must be stupid to believe that reason."
Seeing high quality justifications for opinions you disagree with or meeting people you respect who disagree with you can both lead to attitude changes and even changes of opinion.
But reading punditry on the Internet? Nope.
> Is Facebook trying to find ways to maximize how much bullshit will you agree with?
Well definitely, because most marketing is bullshit. But I don't think that's what this paper is about.
Here's a PDF version http://cn.cnstudiodev.com/uploads/document_attachment/attach...
Thank you! Facebook is blocked where I work, so this gives me a chance to read this when I have some time.
I am convinced (although I do not yet have enough data to prove it) that continuous exposure to a lack of ideologically-diverse information is actually very long-term-bad, leading to unnecessary deaths (i.e., war, people killing each other over mere worldview differences, media taking advantage of polarizing forces to whip people up and generate ad views, etc.).
I base this only on my Psych-major training and just in thinking about things and observing.
If anyone had any actual evidence that might support this conclusion, I would be grateful.
If Facebook actually understands that it is kind of their responsibility at this point to ensure people's feeds are more diverse than they might otherwise prefer (i.e., gently introducing discomfort), this might mark the first time I'm actually in favor of them keeping their feed ranking algorithm secret (so as to quiet the masses, most of whom seem content to never experience ideological discomfort). But this, of course, depends on trusting Facebook...
I base this only on my Psych-major training and just in thinking about things and observing.
If anyone had any actual evidence that might support this conclusion, I would be grateful.
If Facebook actually understands that it is kind of their responsibility at this point to ensure people's feeds are more diverse than they might otherwise prefer (i.e., gently introducing discomfort), this might mark the first time I'm actually in favor of them keeping their feed ranking algorithm secret (so as to quiet the masses, most of whom seem content to never experience ideological discomfort). But this, of course, depends on trusting Facebook...
> I am convinced (although I do not yet have enough data to prove it) ... If anyone had any actual evidence that might support this conclusion, I would be grateful.
This is the opposite of how useful research and/or data science works. Data should be taken as is and then learned from. It certainly should not be gathered in an effort to directly prove a conclusion that you are already "convinced" of.
It's very disheartening to see this as a comment on a data science article on hacker news... unless you're being sarcastic/ironic? This has to be sarcastic/ironic, right? Right? :(
This is the opposite of how useful research and/or data science works. Data should be taken as is and then learned from. It certainly should not be gathered in an effort to directly prove a conclusion that you are already "convinced" of.
It's very disheartening to see this as a comment on a data science article on hacker news... unless you're being sarcastic/ironic? This has to be sarcastic/ironic, right? Right? :(
On the other hand, here we get special insight into the attitude/understanding of statistics common in university psychology programs, and hence a particularly damning look into the field itself. This is how the research works; you come up with a 'just so' gut feeling, and you look and look and look until invariably you come up with /some/ evidence for it and then publish.
> This is the opposite of how useful research and/or data science works. Data should be taken as is and then learned from. It certainly should not be gathered in an effort to directly prove a conclusion that you are already "convinced" of.
Get down off the pulpit. If the data said the opposite, I would instantly change my worldview.
But you are correct, in that my wording or attitude was incorrect. I should have said "I suspect" instead of "I am convinced" and instead of "might support this conclusion" I should have said "might support or refute this hypothesis".
Which is to say, don't scientists at least have a hypothesis in mind before they collect data related to it? Otherwise, why would you be testing at all, and for what exactly? You can't just take millions of data points, put them into a blender and get proven theories out of it!
Anyway, that's what I'd have at this point, a hypothesis. I should have used that wording, my bad.
Get down off the pulpit. If the data said the opposite, I would instantly change my worldview.
But you are correct, in that my wording or attitude was incorrect. I should have said "I suspect" instead of "I am convinced" and instead of "might support this conclusion" I should have said "might support or refute this hypothesis".
Which is to say, don't scientists at least have a hypothesis in mind before they collect data related to it? Otherwise, why would you be testing at all, and for what exactly? You can't just take millions of data points, put them into a blender and get proven theories out of it!
Anyway, that's what I'd have at this point, a hypothesis. I should have used that wording, my bad.
And something that you suspect (or even of which you are convinced) but can't prove is a decent starting place for trying to find the data to prove or refute it.
Evidence doesn't fall from the sky. People create a hypothesis and they seek evidence to test it. Also, your snark is useless.
That's not really true -- there are multiple ways of approaching this.
One camp says "collect any data that might be relevant, and then begin looking at the data to try to figure out what the hypotheses should be"
The other camp says "formulate a hypothesis, and then find the data you need to test that hypothesis".
The problem with the latter approach in the social sciences -- or any setting with lots of unknown latent variables -- is that it's often possible to find some data set for which a given hypothesis holds with p < 0.05. So whenever there are a lot of latent variables, it makes a lot more sense to construct a high quality data set first, and then start hypothesis testing.
The problem with the former approach is that you really need to know "this set of data is probably really interesting / representative for an entire range of hypotheses about topic X", but that's often not clear from the outset. And it's often the case that for any particular hypothesis, there are lots of other data sets you might know could also be relevant.
In any case, whenever there are lots of unknown latent variables, cherry-picking data sets that confirm your hypothesis is a really good way to lead yourself astray.
My solution is to just avoid working in fields with lots of latent variables, but that has limitations was well :-)
One camp says "collect any data that might be relevant, and then begin looking at the data to try to figure out what the hypotheses should be"
The other camp says "formulate a hypothesis, and then find the data you need to test that hypothesis".
The problem with the latter approach in the social sciences -- or any setting with lots of unknown latent variables -- is that it's often possible to find some data set for which a given hypothesis holds with p < 0.05. So whenever there are a lot of latent variables, it makes a lot more sense to construct a high quality data set first, and then start hypothesis testing.
The problem with the former approach is that you really need to know "this set of data is probably really interesting / representative for an entire range of hypotheses about topic X", but that's often not clear from the outset. And it's often the case that for any particular hypothesis, there are lots of other data sets you might know could also be relevant.
In any case, whenever there are lots of unknown latent variables, cherry-picking data sets that confirm your hypothesis is a really good way to lead yourself astray.
My solution is to just avoid working in fields with lots of latent variables, but that has limitations was well :-)
Most interesting quote to me:
"Individual choice has a larger role in limiting exposure to ideologically cross cutting content: after adjusting for the effect of position (the click rate on a link is negatively correlated with its position in the News Feed; see fig. S5), we estimate the factor decrease in the likelihood that an individual clicks on a cross-cutting article relative to the proportion available in News Feed to be 17% for conservatives and 6% for liberals, a pattern consistent with prior research (4, 17). "
"Individual choice has a larger role in limiting exposure to ideologically cross cutting content: after adjusting for the effect of position (the click rate on a link is negatively correlated with its position in the News Feed; see fig. S5), we estimate the factor decrease in the likelihood that an individual clicks on a cross-cutting article relative to the proportion available in News Feed to be 17% for conservatives and 6% for liberals, a pattern consistent with prior research (4, 17). "
(Disclaimer: I didn't read the paper. I'm not sure if someone knows a link is cross-cutting before clicking it.)
So if I'm reading that correctly, liberals are more likely to click on a cross-cutting article. That alone doesn't really tell me much. So without trying to inject too much personal bias...
Does that mean:
1) Liberals are more likely to click on something they disagree with.
So if I'm reading that correctly, liberals are more likely to click on a cross-cutting article. That alone doesn't really tell me much. So without trying to inject too much personal bias...
Does that mean:
1) Liberals are more likely to click on something they disagree with.
a) Because they are more open minded, they click on articles from contrarian points of view?
b) Because the article just looks interesting, but it turns out they disagree with it because they are more closed minded?
a) Because they are more open minded, they agree with more of the things they read than liberals?
b) Because they are more close minded, they don't even bother opening articles they think they will disagree with?This is pure conjecture but...
Liberals in general could be considered the progressives while conservatives are traditionalists. So conservatives are trying to hold onto the old ways (or at least a perversion of them), whereas liberals are looking for reasons why the old ways are wrong and we need to progress.
So conservatives love reading articles that glorify old traditions/ideals, while liberals read the same article in order to confirm all that's wrong with the old ways therefore giving support to their belief that we need a new way to progress forward.
Just my initial speculation, take with a large grain of salt.
Liberals in general could be considered the progressives while conservatives are traditionalists. So conservatives are trying to hold onto the old ways (or at least a perversion of them), whereas liberals are looking for reasons why the old ways are wrong and we need to progress.
So conservatives love reading articles that glorify old traditions/ideals, while liberals read the same article in order to confirm all that's wrong with the old ways therefore giving support to their belief that we need a new way to progress forward.
Just my initial speculation, take with a large grain of salt.
I like this idea as another possible interpretation of that data point. Essentially you're saying that it's because everyone's clicking the more conservative links which causes the skew, albeit for different reasons.
The only point I'm trying to make is that the simple sounding sentence "we estimate the factor decrease in the likelihood that an individual clicks on a cross-cutting article relative to the proportion available in News Feed to be 17% for conservatives and 6% for liberals" really let's you draw your own conclusions.
The only point I'm trying to make is that the simple sounding sentence "we estimate the factor decrease in the likelihood that an individual clicks on a cross-cutting article relative to the proportion available in News Feed to be 17% for conservatives and 6% for liberals" really let's you draw your own conclusions.
Note that the paper observes:
"liberals tend to be connected to fewer friends who share information from the other side, compared to their conservative counterparts: 24% of the hard content shared by liberals’ friends are cross-cutting, compared to 35% for conservatives"
The more articles that I can't agree with I find in my feed, the less likely I am to click on one of them than a guy who sees them less often.
Even if we both read exactly 3 such articles a week.
"liberals tend to be connected to fewer friends who share information from the other side, compared to their conservative counterparts: 24% of the hard content shared by liberals’ friends are cross-cutting, compared to 35% for conservatives"
The more articles that I can't agree with I find in my feed, the less likely I am to click on one of them than a guy who sees them less often.
Even if we both read exactly 3 such articles a week.
I have another conjecture. I think liberals are people who have a very inclusive ingroup, potentially including everybody on Earth; while conservatives have a smaller or better defined ingroup (e.g. Americans).
From this follows that there is only one type of liberals, but many types of conservatives (depending on their ingroup). So the view of liberals and conservatives to the others is different, because conservatives tend to perceive liberals as just another group that has them as outgroup, which is confusing to liberals.
I think in fact it's similar with religion. Atheism is often understood by religious people as another form of religion, but it isn't in fact, it has a unique position among all the types of religious belief; like liberalism is just an edge case of all possible conservativisms.
From this follows that there is only one type of liberals, but many types of conservatives (depending on their ingroup). So the view of liberals and conservatives to the others is different, because conservatives tend to perceive liberals as just another group that has them as outgroup, which is confusing to liberals.
I think in fact it's similar with religion. Atheism is often understood by religious people as another form of religion, but it isn't in fact, it has a unique position among all the types of religious belief; like liberalism is just an edge case of all possible conservativisms.
The classification is binary, so there are no more or less open minded liberals or conservatives. “Neutral” sources are excluded (and rare). You could have an effect of what either side choses to share, say, blue links are outrageous; red one are more mature – but measuring that objectively would be difficult.
Also, Liberals are the one who click less (i.e. “are open minded”) than Conservatives, not the other way around.
That result is a classic: I remember (but can't find the reference) more liberals say they could not marry (or let their child marry) a conservative than the other way around, for instance.
Also, Liberals are the one who click less (i.e. “are open minded”) than Conservatives, not the other way around.
That result is a classic: I remember (but can't find the reference) more liberals say they could not marry (or let their child marry) a conservative than the other way around, for instance.
I would take this with a grain of salt, facebook has an interest to present itself as being unable to effect it's users' opinions.
Is the actual paper hidden behind a login? o.O
Maybe you have to login so they can adapt the paper's conclusions to match your social graph.
It's called "cognitive bubble"
Does it really matter whether people read bullshit that they disagree with? Reading bullshit from either side of the political spectrum isn't really teaching anyone anything (and, moreover, isn't designed to), so why the premium on consuming all varieties of bullshit? The most likely reaction to reading bullshit you disagree with is "This Is Bullshit". Which isn't so surprising when the content really is bullshit.
For me, the more interesting question is whether people are exposed to high-quality justifications for a range of opinions. And I can't remember the last time a research paper showed up on my facebook feed.