When you say “total population”, do you mean total of each race (each bar represents 100% of the female population of the given race), or total overall (each bar represents 100% of the entire female population across races)? Either way, I can see the problem with it, but before going into detail, I want to know what you’re represeting here.
Each bar represents the total population of that race. Assume that the scale on the left would be in percentages rather than absolute numbers.
Lol. I do understand it. And if you understood Bayesian reasoning, you would understand why the two statements are not necessarily contradictory.
Because my assumption is that “all other things are equal.”
Whatever. God forbid you should look at an analysis that might make you realize that you are in the wrong.
You can’t make that assumption, either. If you wish to tout Bayesian probability analysis, it behooves you to apply the careful examination and reassessment of assumptions that is as much the hallmark of that discipline as is the ability to incorporate new data.
The only reason you come up with an answer is because of assumptions you make that would not be acceptable as givens. The analysis is predicated on these faulty assumptions and therefore worthless.
If you don’t wish to make worthless analyses don’t assume operable data.
If a black woman says she is raped, and she is just as honest as a white woman, how can you say the black woman’s claim is more likely to be false. This is the conclusion that makes android’s self-destruct.
If we accept your math as accurate, your conclusion should actually be that if we look at all w-o-b reports and randomly pick one, the chances of picking a false one would be higher than if we did the same thing with b-o-w reports.
Which is no different than predicting that if we roll a dice with 5 blue sides and a red, we have a 83% chance of getting blue and then contrasting that with dice with 3 blues sides and 3 red.
In other words, to suggest that any given black woman’s credibility is less than a white woman simply because there are relatively fewer w-o-b rapes is illogical. Just like a creationist’s, you are using reverse probability to assess an event after it has happened.
Nonsense. The calculation I did simply illustrates that background statistics matter. Anyway, it’s hardly unreasonable to assume, in the absence of special knowledge, that people of different races will behave similarly.
Regardless, why don’t you re-do your math now that you see how you screwed up?
Or are you stuck in “we don’t really know anything for sure” mode?
Because that’s the intellectual equivalent of locking yourself in the bathroom.
My opinion only.
Easily.
I’m not going to bother responding to your arguments. The problem is that you do not understand Bayesian reasoning. If you understood Bayesian reasoning, you would understand that the two statements are not necessarily contradictory.
That’s what I thought. In order for that to hold, we have to make assumptions that weren’t laid down in the premise. We have to assume that blacks in our hypothetical represent a greater percentage of the population than they do as a percentage of rape victims. If blacks represent 1% of rape victims, and also 1% of the population for the area we’re measuring, his argument falls apart. His argument would only hold if blacks represented a much greater percentage of the population than 1% in our hypothetical.
It seems that psychloan was attempting to apply real world ratios of race to the hypothetical even though the numbers given weren’t representative of the real world.
The statements are inherently contradictory, you idiot. There’s no way you can say a black woman is more likely to lie than a white woman, but then hold that they are equally honest.
Lol. By the way, I posted my example in the general questions forum with a few superficial changes to make it non-inflammatory. A few other “idiots” came along and confirmed my calculations.
Except that white woman will be raped 99 times more than black. Is it just me, or it that an odd premise for a world where everything is equal?
Yet you do exactly that. By making bad assumptions you phrase the problem such that 50% of the accusations are made by a group that represents only 1% of the population of rapes.
It’s this kind of garbage in and garbage out that invalidates an analysis.
I haven’t screwed up.
Not knowing, doesn’t mean you can make it up. Probability analysis can’t be based on made up assumptions if it is to be valid. Figuring out whether you really know what you think you know, and working with those things that your sure of is the only way to do a valid analysis.
Guesswork isn’t math.
Yeah, a great big whopping two people, honey chile.
Haha. Go take a statistics class that covers Bayes Thereom. You will quickly learn that the professor is an “idiot,” and the textbook is written by an “idiot,” and the students who pass the class are all “idiots.”
Yes you did. Or have you forgotten already about the cite you demanded and I provided?
Anyway, WHY WON’T YOU REDO YOUR CALCULATION USING 99% and 1% for ACTUAL RAPES?
Why, all of a sudden, do you need all kinds of unknown information?
Eh? No, what you have to assume is that blacks are raped at a substantially lower rate in proportion to population than whites are. That’s what makes the whole thing work. But that’s been a stated assumption from the start. I have no idea if it’s accurate, and I don’t much care. As I’ve said, I don’t think the hypothetical maps to reality in any interesting way for other reasons. All I’ve been trying to do in this thread is clarify how it can be possible for both groups to exhibit the same level of honesty and it still be the case that the reports of one group are much more likely to be false than those of the other.
No, they are actually raped at a higher rate, which is why the hypothetical won’t work if someone wants to try and throw in actual numbers and tries to map them to the hypothetical. What is at a lower rate is blacks raped specifically by whites, but overall, blacks are far more likely to be rape victims per capita than whites.
Well yes, but for the purposes of the hypothetical intraracial rapes were being ignored. Or something. Anyways, I’ve now said at least 3 times that I don’t think the hypothetical maps onto reality. I really don’t care if the assumptions are flawed. Some posters in the thread seemed to be having difficulty with how the conclusion could be drawn even granted the assumptions, and that’s the only point I’ve tried to address.
I’m gonna jump in and overstep my bounds by saying that bitch with the face is off her rocker, and while she may be “smart” she’s a certified idiot.