I think in all those circumstances, the delicacy of the delivery, disavowance of unearned authority, and presence of actionable advice makes the difference between "appropriate" and "inappropriate". I would not be in the slightest bit offended if someone at a party said to me "Listen, I'm no dermatologist, but that mole you said wasn't there last month looks a hell of a lot like one my co-worker had, and it turned out he had skin cancer. Couldn't hurt to get it checked out." Broadly, that tone was what I was trying to achieve.
While I don't disagree there's a negative effect from a "false positive", I think it's likely very small. The worst you'll likely do is offend someone. And while your Pascal's wager analogy is astute, I don't think the same objections quite apply - this case is a much more straightforward one of "high probability of very low harm, vs low probability of very high good", closer to buying a lottery ticket than Pascal's dubious infinities. Precisely calculating the expected return isn't possible, so you have to apply your best estimates.
My post did in fact contain actionable advice - "call your loved ones" - which has the useful benefit of working for many psychological issues besides schizophrenia, and also just being a nice thing to do generally. It was the best advice I could come up with. However, I can see now that in this instance I could probably have achieved the same result by giving the advice without the "diagnosis".
I think in all those circumstances, the delicacy of the delivery, disavowance of unearned authority, and presence of actionable advice makes the difference between "appropriate" and "inappropriate". I would not be in the slightest bit offended if someone at a party said to me "Listen, I'm no dermatologist, but that mole you said wasn't there last month looks a hell of a lot like one my co-worker had, and it turned out he had skin cancer. Couldn't hurt to get it checked out." Broadly, that tone was what I was trying to achieve.
While I don't disagree there's a negative effect from a "false positive", I think it's likely very small. The worst you'll likely do is offend someone. And while your Pascal's wager analogy is astute, I don't think the same objections quite apply - this case is a much more straightforward one of "high probability of very low harm, vs low probability of very high good", closer to buying a lottery ticket than Pascal's dubious infinities. Precisely calculating the expected return isn't possible, so you have to apply your best estimates.
My post did in fact contain actionable advice - "call your loved ones" - which has the useful benefit of working for many psychological issues besides schizophrenia, and also just being a nice thing to do generally. It was the best advice I could come up with. However, I can see now that in this instance I could probably have achieved the same result by giving the advice without the "diagnosis".