I find it likely that the percentage of rapists on LessWrong is roughly comparable to the percentage of rapists in U.S. colleges. Maybe a little lower, maybe a little higher, but not significantly different.
Given a base rate of 6%, I’d be astounded if the rate among male Less Wrong commenters were lower than 3% or higher than 8%; and I would dismiss out of hand a claim that it was lower than 1% or higher than 10%.
In general, I expect that rapists (as we’re using the term here) are present in any large group, and that I have no way of distinguishing them from non-rapists.
It’s to the benefit of women and normal men to develop accurate heuristics to distinguish rapist men from normal men. The “Schrödinger’s Rapist” situation results from such heuristics being absent, or unavailable due to lack of information.
(Yes, I feel okay saying that the 94% of men who are not rapists are “normal men” … and that rapists are not.)
One of the bigger heuristics suggested by the Lisak & Miller study is that repeat rapists commit (on average) about ten times as many non-rape violent crimes as normal men do.
Some other studies suggest other heuristics: rapists have more anger and hostility toward women than normal men do, and rapists have less empathy toward women who have been sexually assaulted than normal men do.
Basically agreed with all of this, though I consider non-male-on-female rape more important than you seem to, which may simply reflect the greater saliency of non-heterosexual relationships to my life more generally.
One of the bigger heuristics suggested by the Lisak & Miller study is that repeat rapists commit (on average) about ten times as many non-rape violent crimes as normal men do.
Which definition of “rapist” was the study using?
Edit: also that reminds me of the argument against acceptance of gays based on statistics showing male homosexuals being ten times more to engage in pedophilia than male heterosexuals.
Here it is. The interesting part is that they ask men whether they have committed particular acts (see the study for which) that legally constitute rape; they don’t ask whether the men think of themselves as rapists.
Edit: also that reminds me of the argument against acceptance of gays based on statistics showing male homosexuals being ten times more to engage in pedophilia than male heterosexuals.
I doubt that claim — and I’m assuming you’re using a folk sense of “pedophilia”, since clinically that term refers to a predilection rather than an act that a person can engage in.
It seems more likely to me that gay sexual relationships which straddle the legal age of consent (in some states, this can mean an 18-year-old boy with a 17-year-old boyfriend) are many, many times more likely to be treated as a criminal issue than straight sexual relationships with the same age gap.
I notice they didn’t bother separating out their data by which of the “rape” questions they answered yes to.
Including the ‘statutory’ kind? If so the study is approximately worthless. (ie. Whatever potential benefit it could have is likely offset by the equivocation it encourages.)
This is by far not the only example of a trait correlated with crime that I suspect you’d rather not act on. I chose homosexuality as the closest analogy since both can be interpreted either as an act or as a predilection and thus a property of the person.
It seems more likely to me that gay sexual relationships which straddle the legal age of consent (in some states, this can mean an 18-year-old boy with a 17-year-old boyfriend) are many, many times more likely to be treated as a criminal issue than straight sexual relationships with the same age gap.
I’m not convinced that’s actually true these days.
Well, the statistic in question is based on data I heard from gay rights advocates. They were saying that only 30% of pedophilia cases are committed by gays and counting on their audience not being Bayesians.
Given a base rate of 6%, I’d be astounded if the rate among male Less Wrong commenters were lower than 3% or higher than 8%; and I would dismiss out of hand a claim that it was lower than 1% or higher than 10%.
It’s to the benefit of women and normal men to develop accurate heuristics to distinguish rapist men from normal men. The “Schrödinger’s Rapist” situation results from such heuristics being absent, or unavailable due to lack of information.
(Yes, I feel okay saying that the 94% of men who are not rapists are “normal men” … and that rapists are not.)
One of the bigger heuristics suggested by the Lisak & Miller study is that repeat rapists commit (on average) about ten times as many non-rape violent crimes as normal men do.
Some other studies suggest other heuristics: rapists have more anger and hostility toward women than normal men do, and rapists have less empathy toward women who have been sexually assaulted than normal men do.
Basically agreed with all of this, though I consider non-male-on-female rape more important than you seem to, which may simply reflect the greater saliency of non-heterosexual relationships to my life more generally.
Which definition of “rapist” was the study using?
Edit: also that reminds me of the argument against acceptance of gays based on statistics showing male homosexuals being ten times more to engage in pedophilia than male heterosexuals.
Here it is. The interesting part is that they ask men whether they have committed particular acts (see the study for which) that legally constitute rape; they don’t ask whether the men think of themselves as rapists.
I doubt that claim — and I’m assuming you’re using a folk sense of “pedophilia”, since clinically that term refers to a predilection rather than an act that a person can engage in.
It seems more likely to me that gay sexual relationships which straddle the legal age of consent (in some states, this can mean an 18-year-old boy with a 17-year-old boyfriend) are many, many times more likely to be treated as a criminal issue than straight sexual relationships with the same age gap.
I notice they didn’t bother separating out their data by which of the “rape” questions they answered yes to.
Including the ‘statutory’ kind? If so the study is approximately worthless. (ie. Whatever potential benefit it could have is likely offset by the equivocation it encourages.)
This is by far not the only example of a trait correlated with crime that I suspect you’d rather not act on. I chose homosexuality as the closest analogy since both can be interpreted either as an act or as a predilection and thus a property of the person.
I’m not convinced that’s actually true these days.
Anecdotes! I have contrary ones.
But what did you think of the Lisak & Miller study, and their definition of rape that you asked about?
That “statistics” is unlikely to be unbiased.
Well, the statistic in question is based on data I heard from gay rights advocates. They were saying that only 30% of pedophilia cases are committed by gays and counting on their audience not being Bayesians.