A touchy subject, I guess

I'm sorry to come across as the dunce here, but could somebody translate those workings into non-mathematical language? I get half way through and I have to start again.

As it stands, I can't seem to understand why it is not possible for there to be more molesters in the priest group and less in the (for example) actors group. If the two groups are the same in total number, a 2% average could be maintained. Am I missing something?

(I expect to shortly feel even more unintelligent...)

Originally posted by Starrman
I'm sorry to come across as the dunce here, but could somebody translate those workings into non-mathematical language? I get half way through and I have to start again.

As it stands, I can't seem to understand why it is not possible for there to be more molesters in the priest group and less in the (for example) actors group. If the two groups are t ...[text shortened]... be maintained. Am I missing something?

(I expect to shortly feel even more unintelligent...)

What's the first step in either proof that you don't follow from the previous?

Originally posted by lucifershammer
This proof is simpler than mine (though mathematically equivalent). Thanks tel.

Your proof by verbosity notwithstanding, why not answer the question?

Originally posted by Starrman
(I expect to shortly feel even more unintelligent...)

ha, you and me both. I get the main idea, that if 2% is the number of molesters in the general population, that 2% could be distributed over any occupation, therefore 2% of any given occupation may be molesters. None of that answers my initial question, though....

Originally posted by David C
Your proof by verbosity notwithstanding, why not answer the question?

What question? Why molesters seek to be priests?

That question only makes sense if the proportion of priests that are molesters were significantly higher than that of the general population. Since that is not the case, your question is meaningless.

My answer would simply be - they are there by [random] chance.

EDIT: It's bad form to call something "proof by verbosity" when you don't understand it. telerion's version of the proof is simpler - if you like I make the same offer to you that Scribs made to Starrman.

Originally posted by lucifershammer


My answer would simply be - they are there by [random] chance.

This of course is not a valid deduction in the same way that your proof is one. It is speculation. There are mathematically consistent alternative explanations.

Originally posted by DoctorScribbles
This of course is not a valid deduction in the same way that your proof is one. It is speculation. There are mathematically consistent alternative explanations.

Of course it's not a valid deduction. But it is [statistically?] as good an explanation as any other.

Originally posted by lucifershammer
No. But it is statistically as good an explanation as any other.

Only in the absence of information that certain occupations attract molesters. If you assume this, then you are assuming what you are trying to demonstrate, namely that there's nothing about the priesthood that appeals to molesters to a greater degree than other occupations.

Suppose we had a population of 200 people, 4 molesters.
One was a teacher. One was a priest. One was a pediatrician. One was a little league coach.

Would you make the same statistical argument, that the molesters are randomly distributed across occupations? Or would you entertain the hypothesis that the trusted access to children that these occupations enjoy might attract molesters more than, say, construction work.

Dr. S

P.S. Actually, we need more people and more molesters, and additional orthogonally arranged occupations to make my example work in the context of the proof.

[response rendered invalid by the later edits of DoctorScribbles]

Originally posted by DoctorScribbles
Only in the absence of information that certain occupations attract molesters. If you assume this, then you are assuming what you are trying to demonstrate, namely that there's nothing about the priesthood that appeals to molesters to a greater degree than other occupations.

Suppose we had a population of 200 people, 4 molesters.
One was a teac ...[text shortened]... ditional orthogonally arranged occupations to make my example work in the context of the proof.

P.S. Actually, we need more people and more molesters, and additional orthogonally arranged occupations to make my example work in the context of the proof.

Not sure what "orthogonal arrangement" is - any help would be greatly appreciated.

But you're right - we would need more people and molesters (hypothetically speaking, of course). In this case, a 90% confidence interval for this population would yield values between 0.4% and 3.6% for the proportion of molesters in the population - making the actual 2% value virtually meaningless.

Originally posted by lucifershammer


Not sure what "orthogonal arrangement" is - any help would be greatly appreciated.

The four suggested categories are similar in several regards, such as their level of:
Authority
Trust
Respect
Expertise
Access to children

Now, take a guy who works at McDonald's, an attorney, a forest ranger and an auto mechanic. Can you construct a similar set of axes on which these all score high? If you can't then I'd call this set of occupations more orthogonal that the original four, using the term in the sense of principal component analysis.

Originally posted by echecero
[response rendered invalid by the later edits of DoctorScribbles]

I don't think so - you raised some good points. Please repost the original if you can.

Originally posted by lucifershammer
I don't think so - you raised some good points. Please repost the original if you can.

I agree. The point of molesters seeking out jobs in the church in the hope of receiving support from the church is one that I hadn't considered.

Originally posted by DoctorScribbles
What's the first step in either proof that you don't follow from the previous?

Once the definitions finish and the maths begins...

How about I set up a theoretical siutaion and you explain to me why mine isn't going to work?

We have a population of 1000
In this population we have 10 occupations with equal numbers of persons (100).
In the occupations we have the following number of molesters:

1) 1
2) 2
3) 3
4) 8
5) 0
6) 1
7) 1
8) 2
9) 1
10) 1

So we have 20 molesters in a population of 1000, which is an average of 2% molester per occupation. But we can plainly see that occupation 4 has several more molesters than the others. How does this compare to the proof that was presented by either Telerion or LH?

Originally posted by Starrman
Once the definitions finish and the maths begins...

How about I set up a theoretical siutaion and you explain to me why mine isn't going to work?

We have a population of 1000
In this population we have 10 occupations with equal numbers of persons (100).
In the occupations we have the following number of molesters:

1) 1
2) 2
3) 3
4) 8
5) 0
6 ...[text shortened]... han the others. How does this compare to the proof that was presented by either Telerion or LH?

Their proofs only speak to occupations (2) and (8), those whose proportions of molesters match the population's.

They conclude that if you were ignorant about all but (2), you could be certain that it wasn't an outlier like (4).