HIV statistics and do people actually understand them.

[trish]

But you wish claim to have proven one third of all gay men carry the HIV virus.

every third gay man has HIV

i.e. you claim to have shown P(H|G) = 1/3, not P(H|G^R) = 1/3. But if we examine your ‘reasoning’ we see you haven’t shown the latter either.

Let’s take the universe W to be R^G; i.e. W=R^G. Then you may replace P(C|R^G) with P(C) as it's already understood (it's given) by your choice of universe one is having an anal receptive encounter with a gay partner. You may also replace P(C|R^G^H) with P(C|H). So using your choice of universe,

P(C) = 0.0027 and P(C|H) = 0.0082.

What you wish to prove is P(H)=1/3. You claim to do this by taking the ratio of the above two quantities. The ratio is

P(C)/P(C|H) = 0.0027/0.0082 = 1/3 (approximately).

By definition of conditional probability P(C|H) = P(C^H)/P(H). So

P(C)/[P(C^H)/P(H)] = 0.0027/0.0082 = 1/3 (approximately).

Simplifying we see that

P(C)P(H)/P(C^H) = 1/3.

This does not reduce to the formula you require.

[trish]

Addendum: We can save your weaker assertion by ignoring the distinction between partners and stipulate C=H. I'm not certain of the legitimacy of such a stipulation, but the result is no longer as surprising as 'one third of all gay men carry the HIV virus.'

[filghy2]

I am only analysing the data given in the tables I gave links to at the beginning of this thread. Y

No, you are propagating a ridiculous conspiracy theory because it happens to fit in with your loopy anti-government fixation. It's nothing but a clever game of starting with the 'answer' and then reverse-engineering dubious calculations to 'prove' it.

You started this thread pretending to be seriously trying to understand these probabilities and people have taken you at your word and gone to some trouble to try and explain it. You are clearly not interested in any explanation that contradicts your conspiracy theory, so I don't intend to waste any more time on you.

[Ts RedVeX]

Can you really not see that in this case your P(C)=P(C^H)? Do you think you can contract HIV from a person that is not infected?! That's why it fucking is 1:3 and the table is a load of crap.

[trish]

But you already agreed that the probability of contracting HIV from a single anal receptive encounter with a gay partner is 0.0027; i.e. you already agreed that P(C) = 0.0027 - now you want it to be equal to 1/3. Yeah it must be the table's fault

You still are long way from proving your claim that: the table implies one out of every three gays has aids.

[trish]

Look, if P(C) = P(C^H), then (in the finite setting of our universe) C=H. In that case, the conditional probability of contracting HIV given a single anal receptive encounter with a gay partner who already carries the virus is P(C|H) = P(C^H)/P(H) = P(C)/P(C) = 1. Your assumption leads to a contradiction without even consulting the table!

[filghy2]

Moreover, if a simple calculation shows that the likelihood of data found on-line being just some sort of homophobic propaganda suggesting that gay men are the only HIV carriers in the world is extremely high, then why bother even looking at the data?

One last very simple question. If what you are suggesting is true then how is it that the probabilities for everything other than gay sex are greater than zero? https://www.aidsmap.com/Estimated-ri...8/#item1324093 They must be very incompetent propagandists, particularly if only a genius like yourself can see the supposed point.

[broncofan]

So she starts a thread titled do people understand hiv statistics. She shows over the course of the thread she doesn't. Then she asserts it's anti-gay propaganda.....apparently all of these organizations are vainly hoping that by making the same errors Redvex did that people will conclude that lots of gay men have hiv. That was the purpose of all that information in the tables. The data on vaginal sex, on oral sex, on anal. It's that if people make a false inference, they will overestimate the prevalence of hiv among gay men.

In the U.S. if you want to know who is homophobic, it's the people who fought to defend the constitutionality of anti-sodomy laws. It's the people who fought tooth and nail to make sure that gay men and women couldn't file joint tax returns. Could not get married. Now that they're married can get snubbed by businesses. You can't make this shit up. But Redvex believes it's the people trying to calculate hiv risk numbers.

[filghy2]

Exactly - if someone wanted to spread homophobic propaganda it is hard to think of a less effective way that relying on people going to this website and performing an obscure calculation. But of course common sense if never a barrier to the dedicated conspiracy theorist.

[trish]

The proper response to post #56 is that the assumption P(C) = P(C^H) only implies that C is a subset of H, and consequently under the assumption P(C|H) is not necessarily unity. I point this out because I like to admit my mistakes. So let’s grant the additional constraint that C is a subset of H within the context that the universe is W = R^G. Then the ratio you’re focused upon is

P(C)/P(C|H) = P(C)/{P(C^H)/P(H)} = P(C)/{P(C)/P(H)} = P(H).

But this is not the probability that a given gay man carries the aids virus as you claim.

That means the probability of encountering an HIV-positive homosexual man (let's assume that is what they mean by "gay") is 1/3 (every third gay man has HIV)....

This is because we’ve agreed at your suggestion to limit the universe to R^G. So P(H) is the conditional probability that a gay man carries the HIV virus given that he has anal receptive sex and is gay.

As you can see, restricting the universe to R^G can be done, but it makes interpretation problematic.

If you see this is the proper interpretation of the ration under the additional constraint, then fine. If not, then let’s go back and take the universe W to be the collection of all possible outcomes. The ratio of interest is then

ratio = P(C|R^G)/P(C|R^G^H).

This, according to the chart is 1/3, but let’s just forget about the chart for now. The denominator is

denominator = P(C|R^G^H).

Your assumption narrower universe was that C is a subset of H. In the broader context this means C^R^G is a subset of R^G^H. So C^R^G^H = C^R^G. Now using the definition of conditional probability


denominator = P(C|R^G^H) = P(C^R^G^H)/P(R^G^H) = P(C^R^G)/P(R^G^H).

Also let’s apply the definition of conditional probability to the numerator of the ratio. Then

numerator = P(C|R^G) = P(C^R^G)/P(R^G).

So now the ratio can be rewritten as

ratio = P(C^R^G)P(R^G^H)/P(R^G)P(C^R^G) = P(H^R^G)/P(R^G) = P(H|R^G).

This is not the promised P(H|G).