I would expect 85% as answer if you believe “In penile-vaginal intercourse with an HIV-infected partner, a woman has an estimated 0.1% chance of being infected, and a man 0.05%. (1 in 2000) Condoms further reduce the risk by 85%” as your link indicates.
Studies that investigated the issue didn’t found 100% protection. I think it’s an interesting test for overconfidence.
HIV transmission rates for any particular sex act are pretty low, even without protection: low enough that the difference from adding a condom usually wouldn’t register at percentage granularity. You’d be better off asking for the transmission rate relative to the unprotected transmission rate, but that’s fairly obscure information; I recall it being considerably higher than I’d anticipated, but I don’t remember exact figures.
In any case, I don’t think this is down to systematic overconfidence so much as it’s a consequence of bad information floating around in the culture. I’ll bet if you asked a hundred people on the street what the chances of contracting HIV from unprotected sex with a carrier were, and then asked them what percentage of that risk they’d experience if they added a condom, they’d overestimate the former and underestimate the latter by very large margins: one or two orders of magnitude on the former wouldn’t surprise me.
P(What’s the likelihood that regular condome use will prevent that you get HIV when you have regular sex with someone with HIV)
Studies that investigated the issue didn’t found 100% protection. I think it’s an interesting test for overconfidence.
What do you mean by “prevent”? P(contagion|no condom)/P(contagion|condom) << 1? If so, does 0.15 count as << 1?
I would expect 85% as answer if you believe “In penile-vaginal intercourse with an HIV-infected partner, a woman has an estimated 0.1% chance of being infected, and a man 0.05%. (1 in 2000) Condoms further reduce the risk by 85%” as your link indicates.
HIV transmission rates for any particular sex act are pretty low, even without protection: low enough that the difference from adding a condom usually wouldn’t register at percentage granularity. You’d be better off asking for the transmission rate relative to the unprotected transmission rate, but that’s fairly obscure information; I recall it being considerably higher than I’d anticipated, but I don’t remember exact figures.
In any case, I don’t think this is down to systematic overconfidence so much as it’s a consequence of bad information floating around in the culture. I’ll bet if you asked a hundred people on the street what the chances of contracting HIV from unprotected sex with a carrier were, and then asked them what percentage of that risk they’d experience if they added a condom, they’d overestimate the former and underestimate the latter by very large margins: one or two orders of magnitude on the former wouldn’t surprise me.