3 ms·

There is nothing else than correlations. How in heaven's name are you going to find a causal relationship? Giving children a placebo phone? > very very dishone

by tgv 3y ago

There is nothing else than correlations. How in heaven's name are you going to find a causal relationship? Giving children a placebo phone? > very very dishonest It isn't dishonest. You know what's dishonest? Calling it "5 minutes of Excel". Jonathan Haidt isn't in the business of selling substack. He's a professor at NYU.

spicymapotofu 3y ago
You might not know that Haidt is a well-known p-hacker and has a record of strongly advising ideology under the guise of social psychology. I haven't trusted any claims of his since his modern peace thesis, and that's not his most outlandish one. Moral psychology appears to overlap 100% with culture war for discourse-killing effects, and Haidt is a giant in the field.
- tgv 3y ago
  I've never heard of Haidt as a p-hacker. A friend of mine is an acquitance of his, which made me assume that Haidt was more in the Bayesian camp. But it's good to view social sciences sceptically. Their theories don't offer much in terms of truth. It's best if they keep their theoretical judgement to themselves, and see where the research takes them. Psychological and sociological theories are not a good basis for policy. The data don't lie, though. Now, there may be issues with the way it was gathered and whathaveyounot, but Haidt isn't exactly the first to observe a large increase in mental health issues. As a matter of fact, the papers have been mentioning it for years by now, as have the health organizations. The correlation from the article can almost not be denied. It's like climate change: you can try to find an alternative explanation (which will bring you into the exact same problems the psychological research you criticize is facing), but until then, we must be cautious. The fact that you don't like messenger is simply not enough to ignore the message.
rtpg 3y ago
In one sense you’re of course right that correlations are all we have. I think the most basic thing is correlating with wealth, or other aspects of a livid environment. “Urban life engenders both early smartphone acquisition and more mental issues (at least self reporting)” is one hypothesis. Another is “social network usage correlates to more mental issues, and having smartphones gives more access to this” (which is relevant! But should be checked. Is it the phone? Is it social networks?) The point being that you are trying to establish a causal narrative, and right now there’s not much to chew on. If we want to help children we should dig deeper to understand actual causes instead of simply reporting the horizontal lines we found.

by tgv 3y ago

spicymapotofu 3y ago

You might not know that Haidt is a well-known p-hacker and has a record of strongly advising ideology under the guise of social psychology. I haven't trusted any claims of his since his modern peace thesis, and that's not his most outlandish one. Moral psychology appears to overlap 100% with culture war for discourse-killing effects, and Haidt is a giant in the field.

tgv 3y ago
I've never heard of Haidt as a p-hacker. A friend of mine is an acquitance of his, which made me assume that Haidt was more in the Bayesian camp. But it's good to view social sciences sceptically. Their theories don't offer much in terms of truth. It's best if they keep their theoretical judgement to themselves, and see where the research takes them. Psychological and sociological theories are not a good basis for policy. The data don't lie, though. Now, there may be issues with the way it was gathered and whathaveyounot, but Haidt isn't exactly the first to observe a large increase in mental health issues. As a matter of fact, the papers have been mentioning it for years by now, as have the health organizations. The correlation from the article can almost not be denied. It's like climate change: you can try to find an alternative explanation (which will bring you into the exact same problems the psychological research you criticize is facing), but until then, we must be cautious. The fact that you don't like messenger is simply not enough to ignore the message.

rtpg 3y ago

In one sense you’re of course right that correlations are all we have. I think the most basic thing is correlating with wealth, or other aspects of a livid environment. “Urban life engenders both early smartphone acquisition and more mental issues (at least self reporting)” is one hypothesis. Another is “social network usage correlates to more mental issues, and having smartphones gives more access to this” (which is relevant! But should be checked. Is it the phone? Is it social networks?) The point being that you are trying to establish a causal narrative, and right now there’s not much to chew on. If we want to help children we should dig deeper to understand actual causes instead of simply reporting the horizontal lines we found.