A scientist realized that the Change My View subreddit is basically a pile of already-structured arguments, with precisely what changed the person’s mind clearly labeled with a “Δ”. He decided to data mine it, and look at what correlated with changed minds.
Conclusions:
Apparently people with longer, more detailed, better structured initial views were more likely to award a delta. (Maybe that’s just because they changed their mind on one of the minor points though, and not the bigger topic? IDK.)
The more people joined the debate, and the longer the comments and comment chains, the more likely the poster was to award a delta. (More and better arguments? I find it interesting that so few people changed their mind on first exposure to someone else’s considerations, and that it takes several back-and-forths to change a view. Maybe they’re going to the meta-level to decide whether it’s valid evidence? Maybe it just takes some dialog to grok the concept? Maybe people are usually aware of the simplest forms of counterarguments to their positions? I’d be interested to know more, but maybe we should be striving for more of a Socratic dialogue. I think there was a recent Eliezer FB post on good arguments being 4+ layers deep, which seems relevant.)
Using the word “we” and insults hurt the odds. (Ingroup/outgroup dynamics, I assume.)
Here, have a related podcast from You Are Not So Smart.
TL;DR:
A scientist realized that the Change My View subreddit is basically a pile of already-structured arguments, with precisely what changed the person’s mind clearly labeled with a “Δ”. He decided to data mine it, and look at what correlated with changed minds.
Conclusions:
Apparently people with longer, more detailed, better structured initial views were more likely to award a delta. (Maybe that’s just because they changed their mind on one of the minor points though, and not the bigger topic? IDK.)
The more people joined the debate, and the longer the comments and comment chains, the more likely the poster was to award a delta. (More and better arguments? I find it interesting that so few people changed their mind on first exposure to someone else’s considerations, and that it takes several back-and-forths to change a view. Maybe they’re going to the meta-level to decide whether it’s valid evidence? Maybe it just takes some dialog to grok the concept? Maybe people are usually aware of the simplest forms of counterarguments to their positions? I’d be interested to know more, but maybe we should be striving for more of a Socratic dialogue. I think there was a recent Eliezer FB post on good arguments being 4+ layers deep, which seems relevant.)
Using the word “we” and insults hurt the odds. (Ingroup/outgroup dynamics, I assume.)