CFLarsen said:
Well, pardon me, then. How else would you describe a person who can unconsciously choose a statistical method that will exhibit bias, before the person has examined the data?
What you believe I meant is irrelevant. You attributed a position to me that I did not or do not hold.
Why do you not want to address this question? This is what you claimed was possible, T'ai Chi.
I did address it and I've said that unconscious bias is natural, not supernatural. That is well known.
Yes, it is.
Bias is only possible when you know what the result is going to be, so you can choose ways to circumvent it.
Yeah, exactly, and they'd know the outcome of using a particular test among other things.
But the issue is not about conscious bias. It is about unconscious bias. Please explain how that works.
I did. In the back of their minds they know which tests are approximations of others hence which would yield a smaller or larger p-value. They might not make a concious choice of which test to use.
No, what you are suggesting is that a statistician can unconsciously choose the method that will show bias, before the data is examined.
Yes, this is easy to do if the statistician has knowledge of the group membership in the analysis and in the interpretation of the analysis. Say you need a p-value smaller than .05 to make the conclusion that coffee is bad for you. Say the evaluator is a coffee fan. He could unconciously select a method of analysis that would give the highest p-value (hence minimizing the interpretation of 'badness'). See the example of Fisher's Exact vs. Chi-square. He could also, in his interpretation of the analysis, unconciously downplay the severity of the judgement against coffee. This is possible, of course, if the evaluator is not blinded.
Not at all. I am focusing on your credentials, as well as on that question you dodge.
"Not at all"? Focusing on a persons' credentials
is focusing on the person and not the issues. I'll play though; you can list
your mathematical and statistical credentials with specific contact information if you want. Personally, I fail to see how it applies to anything. I could have 50 Ph.D's and that would still be irrelevant.
Here's an inequality for you to memorize:
d + L < i,
where d is the number of degrees a person in the discussion has, L is the number of lists of questions you ask, and i is the actual relevant issues, in this case mathematical and statistical, in the discussion.
You can bring up as many d and L as you'd like; that still won't ever get you to i, Claus.
It is not a "game". It is about you answering questions derived from your claims.
Uh huh. Then address some mathematics and statistics then, please. It is about you trying to address those.
Please point it out. Don't just say "some pages", be precise. Please.
I'm not obligated to hold your hand in this. Scroll back and find it yourself.
Up a bit. A little more. There! You got it!
So, you are now saying that the statistician's work should be checked by independent statisticians? Just yes or no, please.
I will never be forced to answer how you desire me to answer.
My answer is that it always depends on specifics. In some experiments, there might be an entire team of evaluators. In others, there might be a sole evaluator. In some experiments, there might be a lengthy peer review process. In others, there might not.
No, I am trying to keep focused here: On your credentials. Why are you so against showing us them?
Irrelevant to the discussion of actual mathematics and statistics. It is just your red herring. I've yanked him out of the water so many times now that he appears to be floating on his back. I'm sure you'll replace him soon...
What?? Your statistical knowledge has been questioned, not just by Hoyt, but also by a professor. It has everything to do with your knowledge of statistics.
More red herrings, as who they are is irrelevant to the issues, just their arguments are important here. You can question away all you want, but sooner or later you'll have to address some actual issues to possibly be taken seriously.
For what its worth as an aside, I've been around Ph.D's for the last 6 years, some of them being my professors and some of them being fellow classmates. I'm not impressed by
just the fact that someone has gone to school a year and a half longer than me, and neither should you be. They have to have good arguments too.
But how?? You just repeat the claim.
And I'll continue flooding you with my answers until hopefully you'll soak it up:
The evaluator is aware of tests, some of which are approximates to other tests and hence yield larger or smaller p-values when they are fed data. This knowledge could be in the back of their minds and could manifest in surprising ways. Not to mention in the interpretation of the analysis.
Yes, that is pretty obvious by now.
Yeah, because you are asking me for specifics on how the brain works. Don't you find anything silly in attempting to guide a person into answering an unanswerable question?
Stop changing the subject, ..
I don't think so Clausypoo. You're not the only one who can ask questions around here, despite your belief otherwise. I'll repeat for you:
Why skeptics are focusing on simply the analysis itself being biased is interesting. Do skeptics ackknowledge unconscious bias can also manifest by an unblinded evaluator in the summary of the results, that is, in the discussion of the analysis?
1) How can you determine how a statistician is able to unconsciously choose the method which will give the biased result?
Done.
2) What are your academic endeavours into the field of statistics? Sources, please.
I mentioned that I have a BS in Math and a MS in Stat, was a teaching assistant for 2 years in low and upper division statistics, a public and private tutor in statistics, have consulted with BS, MS, and Ph.D students on their theses (ranging from biology, zoology plant pathology, and others), and have done and am doing work in areas that are data intenstive, for lack of better words (notice these are not extraordinary claims). All of this is irrelevant to the issues I've raised of course, but since you asked, here you go. Please don't be go Billy on me here and now demand specific personal contacts so you can verify this non-relevant information.
Long ago when I asked Bill if he'd like me to post my degrees and he post his, I didn't hear anything about that, other than degrees are irrelevant, so I left it at that. Now when you and he fail to address the mathematical and statistical issues you want me to post my credentials. I'll repeat:
You can address the actual mathematical and statistical issues if you'd like.
Please either:
- address the question, providing either a retraction or evidence of your claim, or
- state that you refuse to answer.
Another list to ignore! WooHoo!!
We await for you to address some mathematics and statistics someday...