I don’t really have a question. You have a hypothesis:
Transcendental meditation practitioners will reduce the crime rate in their cities in a nonlinear fashion satisfying certain identities.
The statement I have written above I agree with, and would therefore normally downvote.
However, you posit specific figures for the reduction of the crime rate. I have no experience with city planning or crime statistics or population figures, and hence have no real basis to judge your more specific claim.
If I disagreed with it on a qualitative level, then I would upvote. If I had any sense of what your numbers meant I might think that they were about right or too high or too low but since I don’t I’m not able to evaluate it.
But not-evaluating because I don’t know how to engage the numbers is different than not-evaluating because I didn’t read it, so I wanted to make the difference clear; since the point of the game is to engage with ideas that may be controversial.
I’m still not sure I understand what you mean, but let me take a shot in the dark:
Out of the variance in crime rate that depends causally on the size of the meditating group, most of that variance depends on whether or not the size of the group is greater than a certain value that I’ll call x. If the meditating group is practicing only TM, then x is equal to 1% of the size of the population to be affected, and if the meditating group is practicing TM-Sidhis, then x is equal to the square root of 1% of the population to be affected.
For example, with a TM-only group in a city of ten thousand people, increasing the size of the group from 85 to 95 meditators should have a relatively small effect on the city’s crime rate, increasing from 95 to 105 should have a relatively large effect, and increasing from 105 to 115 should have a relatively small effect.
Edit: Or did you mean my confidence values? The second proposition (about the nonlinear relationship) I assign 20% confidence conditional on the truth of the first proposition. Since I assign the first proposition 75% confidence, and since the second proposition essentially implies the first, it follows that the second proposition receives a confidence of (0.2 * 0.75)=15%.
I understand what you meant by your proposition, I’m not trying to ask for clarification.
I assume you have some model of TM-practitioner behavior or social networking or something which justifies your idea that there is such a threshold in that place.
I do not have any models of: how TM is practiced, and by whom; how much TM effects someone’s behavior, and consequently the behavior of those they interact with; how much priming effects like signs or posters for TM groups or instruction have on the general populace; how much the spread of TM practitioners increases the spread of advertisement.
I would not be hugely surprised if it were the case that, given 1% of the population practiced TM, this produced enough advertisement to reach nearly all of the population (i.e. a sign on the side of a couple well-traveled highways) or enough social connections that everyone in a city was within one or two degrees of separation of a TM practitioner.
But I also wouldn’t be surprised if the threshold was 5%, or .1%, or if there was no threshold, or if there was a threshold in rural areas but not urban areas, or conservative-leaning areas but not liberal-leaning areas, or the reverse. I have no model of how these things would go about, so I don’t feel comfortable agreeing or disagreeing.
Certainly fewer than 15% of the possible functions of TM-practice vs crime are as you describe, but it is certainly far more likely that your hypothesis is true compared with the hypothesis “even one TM-practitioner makes the crime rate 100%” but I don’t know if it’s 5 bits more relevant or 10 bits more relevant, and I don’t know what my probabilities should be even if I knew how many bits of credence I should give a hypothesis.
If you know something more than I do (which is to say, anything at all) about social networking, advertising, or the meditation itself, or the people who practice it, then you might reasonably have a good hypothesis. But I don’t, so I can only take the outside view, which tells me “more people actively relaxing/being mindful/spending their time on non-crime activity should reduce crime.”
The causally primary reason for my belief is that while I was growing up in a TM-practicing community, I was told repeatedly that there were many scientific studies published in respectable journals demonstrating this effect, and the “square root of one percent” was a specific point of doctrine.
I’ve had some trouble finding the articles in question on academically respectable, non-paywalled sites (though I didn’t try for more than five or ten minutes), but a non-neutrally-hosted bibliography-ish thing is here.
(Is there a general lack of non-paywalled academically respectable online archives of scientific papers?)
.
(Edited to add: if anyone decides to click any of the videos on that page, rather than just following text links, I’d assign Fred Travis the highest probability of saying anything worth hearing.)
.
(Edited again: I was going to say this when I first wrote this comment, but forgot: The obvious control would be against other meditation techniques. I don’t think there are studies with this specific control on the particular effect in my top-level comment, but there are such studies on e.g. medical benefits.)
.
(Edited yet again: I’ve now actually watched the videos in question.
The unlabeled video at the top (John Hagelin) is a lay-level overview of studies that you can read for yourself through text links. (That is, you can read the studies, not the overview.)
Gary Kaplan is philosophizing with little to no substance in the sense of expectation-constraint, and conditional on the underlying phenomena being real his explanation is probably about as wrong as, say, quantum decoherence.
Nancy Lonsdorf is arguing rhetorically for ideas whose truth is almost entirely dependent on the validity of the studies in question and that follow from such validity in a trivial and straightforward fashion. Some people might need what she’s saying pointed out to them, but probably not the readers of Less Wrong.
Fred Travis goes into more crunchy detail, about fewer studies, than any of the others, but still not as much detail as just reading the papers.)
I don’t really have a question. You have a hypothesis:
Transcendental meditation practitioners will reduce the crime rate in their cities in a nonlinear fashion satisfying certain identities.
The statement I have written above I agree with, and would therefore normally downvote.
However, you posit specific figures for the reduction of the crime rate. I have no experience with city planning or crime statistics or population figures, and hence have no real basis to judge your more specific claim.
If I disagreed with it on a qualitative level, then I would upvote. If I had any sense of what your numbers meant I might think that they were about right or too high or too low but since I don’t I’m not able to evaluate it.
But not-evaluating because I don’t know how to engage the numbers is different than not-evaluating because I didn’t read it, so I wanted to make the difference clear; since the point of the game is to engage with ideas that may be controversial.
I’m still not sure I understand what you mean, but let me take a shot in the dark:
Out of the variance in crime rate that depends causally on the size of the meditating group, most of that variance depends on whether or not the size of the group is greater than a certain value that I’ll call x. If the meditating group is practicing only TM, then x is equal to 1% of the size of the population to be affected, and if the meditating group is practicing TM-Sidhis, then x is equal to the square root of 1% of the population to be affected.
For example, with a TM-only group in a city of ten thousand people, increasing the size of the group from 85 to 95 meditators should have a relatively small effect on the city’s crime rate, increasing from 95 to 105 should have a relatively large effect, and increasing from 105 to 115 should have a relatively small effect.
Edit: Or did you mean my confidence values? The second proposition (about the nonlinear relationship) I assign 20% confidence conditional on the truth of the first proposition. Since I assign the first proposition 75% confidence, and since the second proposition essentially implies the first, it follows that the second proposition receives a confidence of (0.2 * 0.75)=15%.
I understand what you meant by your proposition, I’m not trying to ask for clarification.
I assume you have some model of TM-practitioner behavior or social networking or something which justifies your idea that there is such a threshold in that place.
I do not have any models of: how TM is practiced, and by whom; how much TM effects someone’s behavior, and consequently the behavior of those they interact with; how much priming effects like signs or posters for TM groups or instruction have on the general populace; how much the spread of TM practitioners increases the spread of advertisement.
I would not be hugely surprised if it were the case that, given 1% of the population practiced TM, this produced enough advertisement to reach nearly all of the population (i.e. a sign on the side of a couple well-traveled highways) or enough social connections that everyone in a city was within one or two degrees of separation of a TM practitioner.
But I also wouldn’t be surprised if the threshold was 5%, or .1%, or if there was no threshold, or if there was a threshold in rural areas but not urban areas, or conservative-leaning areas but not liberal-leaning areas, or the reverse. I have no model of how these things would go about, so I don’t feel comfortable agreeing or disagreeing.
Certainly fewer than 15% of the possible functions of TM-practice vs crime are as you describe, but it is certainly far more likely that your hypothesis is true compared with the hypothesis “even one TM-practitioner makes the crime rate 100%” but I don’t know if it’s 5 bits more relevant or 10 bits more relevant, and I don’t know what my probabilities should be even if I knew how many bits of credence I should give a hypothesis.
If you know something more than I do (which is to say, anything at all) about social networking, advertising, or the meditation itself, or the people who practice it, then you might reasonably have a good hypothesis. But I don’t, so I can only take the outside view, which tells me “more people actively relaxing/being mindful/spending their time on non-crime activity should reduce crime.”
I understand now.
The causally primary reason for my belief is that while I was growing up in a TM-practicing community, I was told repeatedly that there were many scientific studies published in respectable journals demonstrating this effect, and the “square root of one percent” was a specific point of doctrine.
I’ve had some trouble finding the articles in question on academically respectable, non-paywalled sites (though I didn’t try for more than five or ten minutes), but a non-neutrally-hosted bibliography-ish thing is here.
(Is there a general lack of non-paywalled academically respectable online archives of scientific papers?)
.
(Edited to add: if anyone decides to click any of the videos on that page, rather than just following text links, I’d assign Fred Travis the highest probability of saying anything worth hearing.)
.
(Edited again: I was going to say this when I first wrote this comment, but forgot: The obvious control would be against other meditation techniques. I don’t think there are studies with this specific control on the particular effect in my top-level comment, but there are such studies on e.g. medical benefits.)
.
(Edited yet again: I’ve now actually watched the videos in question.
The unlabeled video at the top (John Hagelin) is a lay-level overview of studies that you can read for yourself through text links. (That is, you can read the studies, not the overview.)
Gary Kaplan is philosophizing with little to no substance in the sense of expectation-constraint, and conditional on the underlying phenomena being real his explanation is probably about as wrong as, say, quantum decoherence.
Nancy Lonsdorf is arguing rhetorically for ideas whose truth is almost entirely dependent on the validity of the studies in question and that follow from such validity in a trivial and straightforward fashion. Some people might need what she’s saying pointed out to them, but probably not the readers of Less Wrong.
Fred Travis goes into more crunchy detail, about fewer studies, than any of the others, but still not as much detail as just reading the papers.)
Wow that was a super in depth response! Thanks, I’ll check it out if I have time.