Why Facts Don’t Change Our Minds (2017) • Hacker News Board

Re: Why Facts Don’t Change Our Minds (2017)

#121

Post by TangoTrotFox » Mon, Oct 01, 2018, 4:15 PM UTC

netcan wrote: post #32

This is a tangent. How much/long does a study like this cost/take? Is it possible/useful to scale this? From reading into some of the replicability discussion, I've gotten curious about social science generally. What's an expirement like this trying to prove/demonstrate. The journalistic narrative likes the easy "X debunked." I imagine researchers have a more nuanced perspective. I guess what im asking is what does t…

The replication rate is social science studies is very poor. And an even larger problem is that the studies themselves do not actually test what they aim to test, often because they simply cannot. For instance do video games cause violence? The only way you could realistically test this would be to have one group somehow prevented from playing video games for their entire life, or at least up to a certain age, and another group obligated to play video games in a similar fashion, and then some control group presumably free to do as they saw fit. And then follow these individuals for decades and see what happens.

That's not really possible. So instead the experiments that are made are toy experiments, but when you're not really testing what you're trying to measure, it becomes impossible to prove anything, and possible to show anything. For instance could you design a toy experiment that might indicate video games cause violence? Of course. Could you design a toy experiment that might indicate video games don't cause violence? Of course. The experiments are meaningless.

As an example of the problem, 'emotional intelligence' in work is all the rage right now. Yet the keystone study that sparked it is really just quite bad. The author had people split off into groups and perform a variety of tasks such as, literally, planning a shopping list. Using some method of determining who made the best shopping list, the author then determined that the groups who had the highest average IQ did not perform the best, whereas their 'emotional intelligence' as measured by an, again literally, "reading the mind in the eyes test" mapped better to performance - so therefore, it's not merit alone that judges performance but some emotional intelligence, at least as measured by "reading the mind in the eyes" that does. That's just broken logic. At the bare minimum IQ != specific task merit. The most logical way to perform that experiment would be to have created teams of those that performed individually best on any given task to work together against those that did not score so well on merit, but did well on the "reading the mind in the eyes test". Of course she did not do this, the bare minimum to even begin approaching this question, since the result would not be what she wanted. And negative results don't get published. Yet now there have been likely hundreds of articles and spin off studies taking that study's unjustified conclusion as a granted.

So no, there is absolutely no big picture progress in the social sciences.

Re: Why Facts Don’t Change Our Minds (2017)

#122

Post by emodendroket » Mon, Oct 01, 2018, 4:16 PM UTC

klmr wrote: post #112

mrfredward wrote: post #105

Earlier quoted context omitted.

Without endorsing, titanix's view of the article, I have to say it's bothersome that all the political examples leaned the same way. It's an article about confirmation bias, and while most examples were neutral, they only ventured into politics when it would confirm the biases of a left-leaning readership.

> it's bothersome that all the political examples leaned the same way. But they don’t: The article also cites the example of vaccine denialism which is much more pervasive on the political left (though they tie it back to Trump). As an aside I agree that it’s somewhat lazy that these kinds of articles always come back to politics. Then again, it’s hard not to discuss the Trump administration when talking about “post-…

That stuff (and associated new-age health mumbo jumbo) has a lot more pull with hard-right guys than you'd think.

Re: Why Facts Don’t Change Our Minds (2017)

#123

Post by 08-15 » Mon, Oct 01, 2018, 4:17 PM UTC

adrianN wrote: post #80

08-15 wrote: post #78

Earlier quoted context omitted.

Generally, because "something some guy said" is not a fact, and the article makes the very same mistake three times. For example, "95% of climate scientists agree" is not a fact about global warming. It's a fact about climate scientists, and at best a rumor about climate change. What would be a fact? The GISS raw data would come close. But James Hansen doesn't release it. No, the data has to first be "cleaned", lest…

I wouldn't know what to do with raw data about climate change and I'm unusually well educated. Just not in climate science. You can't expect a majority of people to educate themselves sufficiently in every relevant topic to make informed decisions. That's why we have scientists who study things for years and years and specialize in interpreting data. The rational thing to do for almost everything is to trust the majo…

> The rational thing to do for almost everything is to trust the majority of the specialists.

Brilliant way to miss the point, which is that rational reasoning also tells you whom to trust. If you have two groups pointing fingers at each other while yelling "Those are not specialists! We are!", you gain no information about who really is. You need something else, and untampered data is something else, even if you can't analyze it yourself.

Re: Why Facts Don’t Change Our Minds (2017)

#124

Post by merpnderp » Mon, Oct 01, 2018, 4:18 PM UTC

izzydata wrote: post #70

They sure change my mind. I'm just not always convinced what people are saying are facts. At best they are the current best explanation and at worst only opinions.

At one end of the spectrum of what constitutes a fact is Richard Feynman's definition, which was anything he'd replicated himself in a lab. He considered everything else he knew to be based on faith (in people, in organizations, in whatever made the claim, but which he held in high enough esteem to actually believe.)

Re: Why Facts Don’t Change Our Minds (2017)

#125

Post by gravy » Mon, Oct 01, 2018, 4:21 PM UTC

klmr wrote: post #112

mrfredward wrote: post #105

Earlier quoted context omitted.

Without endorsing, titanix's view of the article, I have to say it's bothersome that all the political examples leaned the same way. It's an article about confirmation bias, and while most examples were neutral, they only ventured into politics when it would confirm the biases of a left-leaning readership.

> it's bothersome that all the political examples leaned the same way. But they don’t: The article also cites the example of vaccine denialism which is much more pervasive on the political left (though they tie it back to Trump). As an aside I agree that it’s somewhat lazy that these kinds of articles always come back to politics. Then again, it’s hard not to discuss the Trump administration when talking about “post-…

Well the article itself doesn't state that it's more pervasive on the left and just makes the link to Trump anyway. So it still actually does the lean the same way.

Re: Why Facts Don’t Change Our Minds (2017)

#126

Post by bilbo0s » Mon, Oct 01, 2018, 4:22 PM UTC

7952 wrote: post #111

bilbo0s wrote: post #91

Earlier quoted context omitted.

It's both. It's the reason people ignore facts. And the reason people change their minds. Just to sharpen the point, think of it this way, ideology can make you "change your mind" and start believing things that have absolutely no basis in fact.

A good example of this is the history of medecine. Perfectly intelligent and educated people believed some really wild things about how infection occurred. It took a long time to convince people that hygiene was neccessary, and that hospitals should prevent contamination between patients. Nowadays we all believe in bacteria and viruses without question, but most of us have never seen either with our own eyes. But ove…

Medicine is still, in fact, full of people who believe things that, 100 years from now, will be likely elicit belly laughs. It's all about the ideology. If the tenets of the ideology tell you you're right, then you're right. (Or as right as you need to be if most people are willing to adhere to that ideology.)

Re: Why Facts Don’t Change Our Minds (2017)

#127

Post by merpnderp » Mon, Oct 01, 2018, 4:22 PM UTC

But what is considered a fact? Richard Feynman only considered a thing a fact if he'd replicated the results himself. Everything else he took on faith.

When the FDA said you should limit fat intake, that was considered a fact and only nuts would disagree, but that is now considered a much trickier statement. When the FDA says the flu vaccine is safe, that is a fact and only nuts would disagree, but that is not considered a questionable statement. But it isn't hard for a nut to use the first to confuse the latter.

Re: Why Facts Don’t Change Our Minds (2017)

#128

Post by Brockenstein » Mon, Oct 01, 2018, 4:23 PM UTC

klmr wrote: post #112

mrfredward wrote: post #105

Earlier quoted context omitted.

Without endorsing, titanix's view of the article, I have to say it's bothersome that all the political examples leaned the same way. It's an article about confirmation bias, and while most examples were neutral, they only ventured into politics when it would confirm the biases of a left-leaning readership.

> it's bothersome that all the political examples leaned the same way. But they don’t: The article also cites the example of vaccine denialism which is much more pervasive on the political left (though they tie it back to Trump). As an aside I agree that it’s somewhat lazy that these kinds of articles always come back to politics. Then again, it’s hard not to discuss the Trump administration when talking about “post-…

>The article also cites the example of vaccine denialism which is much more pervasive on the political left

Is it? Or is that how you perceive it? Because vaccine denialism as far as I've seen doesn't abide by political alignment. All it really takes is a poor understanding of science and history and living in a world where modern medicine has been so successful that vaccines are a victim of their own success. Where a lot of the people alive today, in the U.S., can't even comprehend even comprehend diseases being a real threat which makes vaccines seem kind of unnecessary.

Throw in some autism scares and any other BS you can manufacture and a plurality of people from all sorts of beliefs and backgrounds who buy into some wonkass bullshit out of fear and ignorance. And no political party has a lock on that demographic...

Re: Why Facts Don’t Change Our Minds (2017)

#129

Post by emodendroket » Mon, Oct 01, 2018, 4:47 PM UTC

Brockenstein wrote: post #128

klmr wrote: post #112

Earlier quoted context omitted.

> it's bothersome that all the political examples leaned the same way. But they don’t: The article also cites the example of vaccine denialism which is much more pervasive on the political left (though they tie it back to Trump). As an aside I agree that it’s somewhat lazy that these kinds of articles always come back to politics. Then again, it’s hard not to discuss the Trump administration when talking about “post-…

>The article also cites the example of vaccine denialism which is much more pervasive on the political left Is it? Or is that how you perceive it? Because vaccine denialism as far as I've seen doesn't abide by political alignment. All it really takes is a poor understanding of science and history and living in a world where modern medicine has been so successful that vaccines are a victim of their own success. Where…

I'd say it maps better to some political tendencies than others; by necessity American political parties are coalitions of a ton of tendencies with plenty of differences among them.

Re: Why Facts Don’t Change Our Minds (2017)

#130

Post by Brockenstein » Mon, Oct 01, 2018, 4:51 PM UTC

gameswithgo wrote: post #53

ryanmonroe wrote: post #50

Earlier quoted context omitted.

"Do humans cause climate change?" is a question of fact. "What type of gun control should we have?" is not a question of fact, so you shouldn't expect that simply stating some facts relevant to the issue would automatically change someone's opinion.

I agree, but there are facts about gun issues that make that gun control question harder to answer. People who believe strongly that guns make a country much safer. People who don't understand what assault rifles are, or that you can shoot up a school 99% as well with a `regular` rifle, people who think lack of guns will make a country MUCH safer. There are plenty of empirical evidence to refuse those kinds of belief…

>People who don't understand what assault rifles are,

Or don't care for pedants who try arguing that an AR-15 isn't an assault rifle... because the issue isn't about terminology. Although trying argue about that does deflect from the real issue though.

>or that you can shoot up a school 99% as well with a `regular` rifle

Are you trying to say a bolt action hunting rifle is 99% as effective at massacring large numbers of people compared to a magazine fed semi-automatic weapon? Since you seem to have some complaint about "assault rifles" being used colloquially you might want to define what is a regular rifle, technically.

>people who think lack of guns will make a country MUCH safer. There are plenty of empirical evidence to refuse those kinds of beliefs but people hold on to them pretty strongly.

What you mean to say is you can tolerate the tens of thousands of gun deaths each year. And your belief is that if we take all those guns away those same number of deaths will just transfer over to stabbings or poisonings. Well while I'm not a badass or anything, I'll take my chances with a knife wielding maniac over a gun wielding maniac.

So... all those developed first world nations that have strong gun control laws still have an equivalent amount of murder and weapon assaults because everyone who would have used a gun still picks up a less dangerous weapon and is somehow just as efficient as they would be with a gun?

The biggest problem the U.S. has with gun control is that there's so many guns in the country that it will take a generation, maybe a 100 years for anything but the most draconian laws to have a tangible effect. And the whole time you'll have 2nd amendment wonks crying every day for 100 years about how gun control doesn't work because it didn't instantly fix the issue. However doing nothing hasn't been working either.