I do like this sort of instrumental rationality post, but would like to see actual citations rather than mere claims of “overwhelming” evidence. Thanks for your service.
I got explicit permission from lukeprog to make this post without having lukeprog levels of citations. He’s set a sort of impossible standard for mortals to follow. I agree that I would also like to see such citations, but I’m busy, and this post took me thousands of hours of procrastination and three hours of actual writing. I will do what I can to come back through and add citations later, or address specific citation wants in the comments.
Of course, I’m not actually in a position to give anyone “permission” on such matters. What I remember saying was that “As far as I can tell I am the only one penalized by not including citations, so hopefully you’ll do just fine. I’d rather see the post written than have it be delayed a month or never happen due to an expectation for citations.” Eliezer rarely gave citations, and people seemed to like his posts just fine. And if LW writers begin to think they need to have a lukeprog quantity of citations, then every LW writer will procrastinate on posts until the Singularity happens.
Anyway, thanks for writing this post. I wish I could afford an Aeron. I may get the Furinno.
People’s expectations of citations are really high compared to people’s willingness to cite.
Like, the citing is aversive/boring enough that it’s currently a major barrier to me continuing to post about consciousness. I would have probably written at least twice as many articles as I did if I didn’t have to cite.
I really love the well-cited content, but think that we’re definitely losing articles that I’d like to read because they think that they should be citing things.
The problem isn’t really lacking citations (after all, Yudkowsky’s posts generally don’t have many citations). The problem is saying “The evidence for X is overwhelming”, while failing to provide any evidence of X. It’s effectively saying “take my word for it”.
I use the word “overwhelming” exactly once, on barefoot running. I realize now that I meant to link to the relevant Wikipedia article, which provides the overwhelming evidence. Updated the post.
I’m just being difficult now, but I don’t think this wikipedia article provides “overwhelming” evidence. I’ve read it and I’m still a barefoot skeptic, particularly since the fad has not been prevalent for long enough for us to really get good evidence that it doesn’t cause as many problems as it solves. In particular, the “citation needed”s on the rebuttals to the criticisms of barefoot makes me nervous about trusting the pro-barefoot evidence, especially when it’s equally likely that it’s just been selectively written (like the vast majority of paleo diet literature—the other odd thing people rave about in the fitness community).
Bravo on the other parts of the article though, I’ve found them enlightening.
I agree that “overwhelming” overstates the strength of the evidence. The main argument in favor of barefoot running is that it leads to fewer injuries than shod running, so for the evidence to be overwhelming it would need to include several studies (by different researchers with various study designs) directly showing this advantage in injury rates. But there seem to be very few studies which empirically compare injury rates for barefoot and shod runners. It’s not clear if any of the studies cited in the Wikipedia article actually do so, and the review article that they cite observes that “well-designed studies of the effects of barefoot and shod running on injury are lacking” (Warburton, 2001). I saw Daniel Lieberman give a talk about his biomechanical research on barefoot running, and I remember that he said much the same thing: there’s a shortage of studies looking at whether barefoot running reduces injury rates.
It would be more accurate to say that there’s a growing body of suggestive evidence pointing to the advantages of barefoot running (including research on the biomechanics of barefoot running and studies comparing injury rates for different types of shoes), or something similar but briefer.
I do like this sort of instrumental rationality post, but would like to see actual citations rather than mere claims of “overwhelming” evidence. Thanks for your service.
I got explicit permission from lukeprog to make this post without having lukeprog levels of citations. He’s set a sort of impossible standard for mortals to follow. I agree that I would also like to see such citations, but I’m busy, and this post took me thousands of hours of procrastination and three hours of actual writing. I will do what I can to come back through and add citations later, or address specific citation wants in the comments.
Of course, I’m not actually in a position to give anyone “permission” on such matters. What I remember saying was that “As far as I can tell I am the only one penalized by not including citations, so hopefully you’ll do just fine. I’d rather see the post written than have it be delayed a month or never happen due to an expectation for citations.” Eliezer rarely gave citations, and people seemed to like his posts just fine. And if LW writers begin to think they need to have a lukeprog quantity of citations, then every LW writer will procrastinate on posts until the Singularity happens.
Anyway, thanks for writing this post. I wish I could afford an Aeron. I may get the Furinno.
People’s expectations of citations are really high compared to people’s willingness to cite.
Like, the citing is aversive/boring enough that it’s currently a major barrier to me continuing to post about consciousness. I would have probably written at least twice as many articles as I did if I didn’t have to cite.
I really love the well-cited content, but think that we’re definitely losing articles that I’d like to read because they think that they should be citing things.
The problem isn’t really lacking citations (after all, Yudkowsky’s posts generally don’t have many citations). The problem is saying “The evidence for X is overwhelming”, while failing to provide any evidence of X. It’s effectively saying “take my word for it”.
I use the word “overwhelming” exactly once, on barefoot running. I realize now that I meant to link to the relevant Wikipedia article, which provides the overwhelming evidence. Updated the post.
I’m just being difficult now, but I don’t think this wikipedia article provides “overwhelming” evidence. I’ve read it and I’m still a barefoot skeptic, particularly since the fad has not been prevalent for long enough for us to really get good evidence that it doesn’t cause as many problems as it solves. In particular, the “citation needed”s on the rebuttals to the criticisms of barefoot makes me nervous about trusting the pro-barefoot evidence, especially when it’s equally likely that it’s just been selectively written (like the vast majority of paleo diet literature—the other odd thing people rave about in the fitness community).
Bravo on the other parts of the article though, I’ve found them enlightening.
I agree that “overwhelming” overstates the strength of the evidence. The main argument in favor of barefoot running is that it leads to fewer injuries than shod running, so for the evidence to be overwhelming it would need to include several studies (by different researchers with various study designs) directly showing this advantage in injury rates. But there seem to be very few studies which empirically compare injury rates for barefoot and shod runners. It’s not clear if any of the studies cited in the Wikipedia article actually do so, and the review article that they cite observes that “well-designed studies of the effects of barefoot and shod running on injury are lacking” (Warburton, 2001). I saw Daniel Lieberman give a talk about his biomechanical research on barefoot running, and I remember that he said much the same thing: there’s a shortage of studies looking at whether barefoot running reduces injury rates.
It would be more accurate to say that there’s a growing body of suggestive evidence pointing to the advantages of barefoot running (including research on the biomechanics of barefoot running and studies comparing injury rates for different types of shoes), or something similar but briefer.