Any difference between men and women on average is just that: on average. Think almost-but-not-quite-completely overlapping Gaussian curves. You have a lot more information about your parents than James_Miller, so he’s making a complete guess based on the slight difference in prior for men and women, whereas you’re able to update on much more complete and relevant information about your actual parents, not just the barely relevant fact that one is a man and one is a woman. Conclusion: discuss it with your mother, if that seems better to you.
Any difference between men and women on average is just that: on average. Think almost-but-not-quite-completely overlapping Gaussian curves.
The second sentence does not follow from the first. It is also possible for the Gaussian curves to be so far apart that there is almost no overlap, and that situation is still perfectly describable by saying that there is a difference between both populations on average but reflects a much stronger difference in prior probability. As a matter of empirical fact, only 20% of Alcor’s members were female as of 1999, and the number of women opposed to their husband’s cryonics arrangements is well-known as the hostile wife phenomenon. Combine with Dr. Miller’s experience and we have a strong outside view case that ilzolende’s mother will probably react worse than her father to cryonics.
No, of course it doesn’t follow automatically, but a lot of the time people point out an average difference between men and women, this is the case. I happen to think it’s quite likely that there are good explanations for the phenomena you cite that don’t include “women are intrinsically more biased against cryonics than men”; there are certainly possible explanations, so it would be a bit daft to assume that that one possibility explains all the variance.
I happen to think it’s quite likely that there are good explanations for the phenomena you cite that don’t include “women are intrinsically more biased against cryonics than men”
I think the explanation is that women are intrinsically more conformist then men and since cryonics is currently unusual and perceived as weird, well.
so it would be a bit daft to assume that that one possibility explains all the variance.
The rule of thumb is that 20% of the causes are responsible for 80% of the effect.
The issue is who can better predict what ilzolende’s mom will think of cryonics, ilzolende who knows his mom, or me who has never met ilzolende’s mom but who has talked to many women about cryonics. The chance of my being a better predictor is increasing in the number of women I have talked to about cryonics so while what I wrote didn’t “contradict” what you said it does reduce the likelihood of your being right.
I’m female, but this site doesn’t have a bio section in the user profile, so no big deal.
My estimated probability that my mom will react negatively to discussion of cryonics is being adjusted upwards slightly, but the information I really need is what sort of things someone who is disgusted by cryonics would also be disgusted by.
I would assume what James Miller is saying is that he has seen a wide variety of women react to cryonics, so in addition to having a reliable estimate of the average woman’s behavior, he would know of more women who are similar to a described woman, and how those women reacted.
I would assume that you have more information about how different types of people react to cryonics, and that I have more information about what mental traits my mother has. I tried to pass on the knowledge of her mental traits which I thought were relevant to your estimate, so that you could make a better estimate. I wasn’t suggesting that you didn’t know how people react to cryonics.
It actually is possible to create a bio section on your user profile, but it requires jumping through some hoops. Specifically, if there’s a page for User:yourname on the wiki, the main site pulls it in as a profile and it becomes your main user page (you can still get to your comments by clicking “Comments” or “Overview”): see for example Gwern’s user page on the main site, and on the wiki.
The chance of my being a better predictor is increasing in the number of women I have talked to about cryonics
Nope. By increasing your sample size you are getting your sample mean and so your estimate of the true mean closer to the population mean. But you can never get better than the population mean in your forecast. Someone who knows how the specifics of a particular data point differ from the average/expected value is quite likely to produce a better forecast.
Any difference between men and women on average is just that: on average. Think almost-but-not-quite-completely overlapping Gaussian curves. You have a lot more information about your parents than James_Miller, so he’s making a complete guess based on the slight difference in prior for men and women, whereas you’re able to update on much more complete and relevant information about your actual parents, not just the barely relevant fact that one is a man and one is a woman. Conclusion: discuss it with your mother, if that seems better to you.
The second sentence does not follow from the first. It is also possible for the Gaussian curves to be so far apart that there is almost no overlap, and that situation is still perfectly describable by saying that there is a difference between both populations on average but reflects a much stronger difference in prior probability. As a matter of empirical fact, only 20% of Alcor’s members were female as of 1999, and the number of women opposed to their husband’s cryonics arrangements is well-known as the hostile wife phenomenon. Combine with Dr. Miller’s experience and we have a strong outside view case that ilzolende’s mother will probably react worse than her father to cryonics.
No, of course it doesn’t follow automatically, but a lot of the time people point out an average difference between men and women, this is the case. I happen to think it’s quite likely that there are good explanations for the phenomena you cite that don’t include “women are intrinsically more biased against cryonics than men”; there are certainly possible explanations, so it would be a bit daft to assume that that one possibility explains all the variance.
I think the explanation is that women are intrinsically more conformist then men and since cryonics is currently unusual and perceived as weird, well.
The rule of thumb is that 20% of the causes are responsible for 80% of the effect.
I have discussed cryonics with hundreds of women. (I talk about it with my students and I teach at a women’s college.)
Sure, I don’t think that contradicts what I said?
The issue is who can better predict what ilzolende’s mom will think of cryonics, ilzolende who knows his mom, or me who has never met ilzolende’s mom but who has talked to many women about cryonics. The chance of my being a better predictor is increasing in the number of women I have talked to about cryonics so while what I wrote didn’t “contradict” what you said it does reduce the likelihood of your being right.
I’m female, but this site doesn’t have a bio section in the user profile, so no big deal.
My estimated probability that my mom will react negatively to discussion of cryonics is being adjusted upwards slightly, but the information I really need is what sort of things someone who is disgusted by cryonics would also be disgusted by.
I would assume what James Miller is saying is that he has seen a wide variety of women react to cryonics, so in addition to having a reliable estimate of the average woman’s behavior, he would know of more women who are similar to a described woman, and how those women reacted.
I would assume that you have more information about how different types of people react to cryonics, and that I have more information about what mental traits my mother has. I tried to pass on the knowledge of her mental traits which I thought were relevant to your estimate, so that you could make a better estimate. I wasn’t suggesting that you didn’t know how people react to cryonics.
Thanks for the data!
It actually is possible to create a bio section on your user profile, but it requires jumping through some hoops. Specifically, if there’s a page for User:yourname on the wiki, the main site pulls it in as a profile and it becomes your main user page (you can still get to your comments by clicking “Comments” or “Overview”): see for example Gwern’s user page on the main site, and on the wiki.
Most people don’t bother, though.
Nope. By increasing your sample size you are getting your sample mean and so your estimate of the true mean closer to the population mean. But you can never get better than the population mean in your forecast. Someone who knows how the specifics of a particular data point differ from the average/expected value is quite likely to produce a better forecast.