Dr. Stewart's DNA Paper...

[juliann]

Yes, this is a good idea. But it should be Fair. How about: you are allowed to discuss if you have under 5,000 posts on this board. Is that too exclusionary?

Gervin's cheap shot is a good example of the methodology that I have been objecting to from the get-go.....

The Dude

Posts: 2852

Joined: 13-November 05

Member No.: 2451

Juliann

Posts: 5313

Joined: 26-November 03

Member No.: 6

I can't keep up with the technical nature of the debate, I'll bet few can. But like the above, everyone can see when something has been omitted, changed or taken out of context.

[The Dude]

Whatever, as in Juliann is absurd.

The Dude is not adding much more than pictures and name-calling.

Obviously the science is way over her head, so she's only able to focus on the polemics. I hope she enjoys the colorful pictures, though.

You have consistently underestimated Stewart, TD.

No, he's exactly what I thought he would be. Exactly.

[Tarski]

You seriously don't get. We are right back where we started... TD's methodology.

Oh brother.

Do you really think you understand the science? Let me remind you that The Dude is defending not only mainstream science but also the position of LDS scientists like Keith Crandall and David McClellan who are actually in a relevant field.

If you think you understand the science (in particular the equations I mentioned) then lets hear it.

If you can't then I suggest you back down.

[Confidential Informant]

Whatever, as in Juliann is absurd.

Obviously the science is way over her head, so she's only able to focus on the polemics. I hope she enjoys the colorful pictures, though.

No, he's exactly what I thought he would be. Exactly.

Really. Well I at least have some ability to follow the science here and, quite frankly, it looks to me like you are getting your *** kicked.

C.I.

[The Dude]

Really. Well I at least have some ability to follow the science here and, quite frankly, it looks to me like you are getting your *** kicked.

C.I.

What's really touching here is the way you guys say nothing, nothing, about my posts, but you all jump for joy when your champion makes a post. I'm sure you know what you are talking about CI, and of course you are a totally neutral judge.

[Jerubaal]

Really. Well I at least have some ability to follow the science here and, quite frankly, it looks to me like you are getting your *** kicked.

C.I.

I'm in the library and you're making me burst out laughing! Quiet!

[Confidential Informant]

What's really touching here is the way you guys say nothing, nothing, about my posts, but you all jump for joy when your champion makes a post. I'm sure you know what you are talking about CI, and of course you are a totally neutral judge.

Juliann's got me. You've got Tarski. It's all good.

C.I.

Edited to add: Dude, it occurs to me that Juliann's point goes directly to the heart of your posts. If you want to make it a purely scientific debate, then why are you interjecting all the mockery and name calling? If your argument is strong make it. But let me tell you something. In court, when the facts are against an attorney and things aren't going well, the very first line of defense is to attack the character of the witnesses. To me, it appears that that is exactly what you are engaged in here.

That's just an observation, take it for what it's worth.

ci (faithful pet yorkie of Juliann)

[juliann]

Oh brother.

Do you really think you understand the science?

Um, no...that is why I said I didn't. But that has nothing to do with the complaints of convenient and unscientific omissions and changes, does it. More examples of how to leave things out so you can change the topic?

[Tarski]

I think everyone seemed rather impressed by Dr. Stewarts post wherein he challenged The Dude's claim the the basic relation for mutation-rate was linear. He is maintaining this even now! I take it part of what seemed so impressive was his splashing up sophisticated looking nonlinear equations such as the following:

However, Dr. Stewart excused himself from discussing the equations. How do those equations really work to undermine The Dude's analysis? We are left guessing. In fact, I think everyone should consider the following:

One can't just throw out a bunch of equations and say that they are nonlinear and expect the point to be made. Here's why:

Look at the following nonlinear equations:

Well, these are the equations for a very basic statistical method known as linear regression. Thats right LINEAR regression. The fact that these equations are nonlinear has nothing to do with the fact that underlying model is linear. The distance from a data point to a line is quadratic. So what? The formulas for standard deviations and the like are nonlinear. So what? Nonlinear equations abound in a statistcal analysis of linear models. Furthermore, just becuase something is nonlinear does not mean that it is exponential! To make his point work, Dr. Stewart would not only have to show that the error (not the underying model) was exponential but also show that the exponential growth is such that the uncertainties for this particular case are as big as he suggested. Even exponential growth in error is OK if the exponential growth rate is low enough and the time short enough. For example, if the error growth was then even after 10,000 years the error would be less than about 10% as long as the number "a" is around 2.3x10^-4.

It seems that Dr. Sterwart just may have been attempting to bamboozle us with his display of unexplained equations.

Mr. Ph.D. Statisitician friend whose office is two doors down from me has looked over the arguments here as a favor and says that he feels that your understanding of the statistics (and questions of error, nonlinearity etc.) is perhaps lacking.

Can you clarify? I think there are many here with enough background to follow the math and see if there really is the kind of error you need to make your points. So far your use of those equations looks like a smoke screen. It reminds me of arguments against evolution that appeal to supposed errors in radiometric dating. All of those arguments are ill considered.

But maybe you can clarify.

[juliann]

Obviously the science is way over her head, so she's only able to focus on the polemics.

I just said that. In fact, that is the point...which seems to have flown right over your head.

I hope she enjoys the colorful pictures, though.

I did, thank you.

No, he's exactly what I thought he would be. Exactly.

People who resort to juvenile name-calling and "email" don't come off nearly as self-assured as you claim you are.

*edited to fix the formatting

[Tarski]

Really. Well I at least have some ability to follow the science here and, quite frankly, it looks to me like you are getting your *** kicked.

C.I.

I just posted a long analysis here and I get a server connection error and then the whole post is lost. Annoying.

My jist was of course that I disagree. Oh well.

[juliann]

I think everyone seemed rather impressed by Dr. Stewarts post wherein he challenged The Dude's claim the the basic relation for mutation-rate was linear.

I'm impressed with both of them when it comes to knowledge. That isn't the point and neither is the science.

This is the point...and the reason I was essentially escorted out of this thread when I said much the same thing when this started out:

My patience has been tried with the need to continually correct your egregious mis-quoting of my statements and of those found in the genetic literature. What is perhaps most concerning is that you seem to be oblivious to the inaccuracies in your own citations and summaries of othersâ?? statements.

No one has to understand the science to be able to see mis-quoting and misrepresentations of the other guy. I saw it before...I see it now. And it appears to me that is why TD is escalating with the insults.

[Nighthawke]

The Dude,

You didn't answer my question. Are you preparing a response, conceding, or perhaps just mulling things over? I don't understand why your response was so incredibly rude. My question was the first time I've ever posted in this thread.

[The Dude]

Edited to add: Dude, it occurs to me that Juliann's point goes directly to the heart of your posts. If you want to make it a purely scientific debate, then why are you interjecting all the mockery and name calling? If your argument is strong make it. But let me tell you something. In court, when the facts are against an attorney and things aren't going well, the very first line of defense is to attack the character of the witnesses. To me, it appears that that is exactly what you are engaged in here.

This is true, CI. After my last Pundit post, the only comment from a Stewart supporter was from Hawkmoon who was disappointed with some belittling comments I made. I willingly made changes because, indeed, I don't need to belittle him to make my real scientific points.

(I did leave in "nincompoop" because I thought it was such an accurate label after his critique of Seielstad. If you really are up to speed on the topics, CI, maybe you should read the Seielstad paper for yourself. Do you want me to email the pdf to you? It's only ~2 pages of text.)

[Confidential Informant]

I just posted a long analysis here and I get a server connection error and then the whole post is lost. Annoying.

My jist was of course that I disagree. Oh well.

Don't worry. It looks to me like the post made it.

C.I.

[The Dude]

The Dude,

You didn't answer my question. Are you preparing a response, conceding, or perhaps just mulling things over? I don't understand why your response was so incredibly rude. My question was the first time I've ever posted in this thread.

What did I say that was so incredibly rude?

I plan to post a short summary of my position. Short. I don't have time to carry on the Pundit thread as I have been, and it seems Stewart doesn't have time for it either. We should both be commended for putting up with this for so long, in front of an audience that will ultimately judge based on rhetorical skill and initial bias instead of the scientific arguments put forth. It's a bummer, man....

[Tarski]

I'm impressed with both of them when it comes to knowledge. That isn't the point and neither is the science.

Ah, but you cannot judge whether something left out had the importance claimed unless you understand the science. Just because Stewart says that something left out was important doesn't mean its true. You are just accepting that assessment without having the background.

What about the huuuuuge ommision I pointed out? He claims significant exponential nonlinearities and his proof was a splash of equations mostly of the type that apply even for linear models. For example, standard deviation and measures of deviation are nonlinear even for linear models (look at any explanation of linear regression). Wasn't his leaving out of that fact a serious omission (or lack of understanding)?

So is the correct model nonlinear or is it just that some nonlinear equations show up as is usual even in linear models?

It could be either way in this case but we are never given the explanation. Why? I think I know why.

[asbestosman]

I think everyone seemed rather impressed by Dr. Stewarts post wherein he challenged The Dude's claim the the basic relation for mutation-rate was linear. He is maintaining this even now! I take it part of what seemed so impressive was his splashing up sophisticated looking nonlinear equations such as the following:

Let me give a stab at the equations. The first one has both sides squared. That's particularly funny because when both sides are squared, one could just as easily remove the square and keep in mind the absolute value. That makes it linear over long stretches if m1 and m2 are changed proportionally. If only one is changed, the equations are not linear in the sense that a proportional increase will appear on the other side but is linear in the sense of being in a line.

The second one also appears linear, but in a multilinear algeraic sort of way. If you only increase the Xi together and proportionally (as though the Xi are components of a vector), then you end up with a linear change in D1. The same argument holds if you do it to Yi while leaving Xi constant. One might write this out in another form:

D1 = yT * M * x

Where yT is a transposed vector. x is a vector, and M is the matrix:

0   1   4   9   16 . . .   i^2
1   0   1   4   9 . . .	(i-1)^2
4  1	0   1   4 . . .	(i-2)^2
. . . 
j^2  (1-j)^2  . . .	   (i-j)^2

And all within the umbrella of linear algebra.

Now this doesn't prove the effects are linear. For all I know x and y could vary together so we'd get quadratic effects (not eponential). On the other hand, the relationships may refer to differential equations in which case exponential results are likely (although it might be exponentially decreasing, or even periodic as a sine wave).

[Tarski]

I'm impressed with both of them when it comes to knowledge. That isn't the point and neither is the science.

Ah, but you cannot judge whether something left out had the importance claimed unless you understand the science. Just because Stewart says that something left out was important doesn't mean its true. You are just accepting that assessment without having the background.

What about the huuuge ommision I pointed out. He claims significant exponential nonlinearities and his proof was a splash of equations mostly of the type that apply even for linear models. For example, standard deviation and measures of deviation are nonlinear even for linear models (look at any explanation of linear regression). Wasn't his leaving outof that fact a serious omission (or lack of understanding)?

So is the correct model nonlinear or is it just that some nonlinear equations show up as is usual even in linear models?

Further, exponetial errors are possible even for linear models so this seems confused all the way around.

It could be either way in this case but we are never given the explanation. Why? I think I know why.

Let me give a stab at the equations. The first one has both sides squared. That's particularly funny because when both sides are squared, one could just as easily remove the square and keep in mind the absolute value. That makes it linear over long stretches except at a crease right where the equations are equal to 0 (m1 = m2).

The second one also appears linear, but in a multilinear algeraic sort of way. If you only increase the Xi together and proportionally (as though the Xi are components of a vector), then you end up with a linear change in D1. The same argument holds if you do it to Yi while leaving Xi constant. One might write this out in another form:

D1 = yT * M * x

Where yT is a transposed vector. x is a vector, and M is the matrix:

0   1   4   9   16 . . .   i^2
1   0   1   4   9 . . .	(i-1)^2
4  1	0   1   4 . . .	(i-2)^2
. . . 
j^2  (1-j)^2  . . .	   (i-j)^2

And all within the umbrella of linear algebra.

Exactly. (And thanks for stepping accross party lines a bit).

[Nighthawke]

What did I say that was so incredibly rude?

I plan to post a short summary of my position.

Well if I thought it was rude then I'm not going to repeat it am I?

Thank you for answering my question.

[juliann]

Ah, but you cannot judge whether something left out had the importance claimed unless you understand the science. Just because Stewart says that something left out was important doesn't mean its true. You are just accepting that assessment without having the background.

I don't have to understand your equations to understand the importance of leaving out a critical qualifier like this:

(Seielstad)

There is now compelling evidence that the male generation times -- at least within modern times â?? is closer to 32 or 35 years.

You [TD] conveniently omitted the crucial phrase â?? â??within modern times.â?

[Tarski]

I don't have to understand your equations to understand the importance of leaving out a critical qualifier like this:

You [TD] conveniently omitted the crucial phrase â?? â??within modern times.â?

And likewise you omit "at least". Without that, it seems like it will mean "only in modern times". But thats not what is said: The "at least within modern times" was mild qualifier- it sounds like it is really believed to be true in the past also, that is, the implication is that it is probably true for ancient times as well.

So why was "at least" left out? A tiny bit of hypocrisy?

[The Dude]

I don't have to understand your equations to understand the importance of leaving out a critical qualifier like this:

You [TD] conveniently omitted the crucial phrase â?? â??within modern times.â?

If Stewart had gone and read the Tremblay paper, which Seielstad cited, then Stewart wouldn't have made the ridiculous assertion that 25 or 35 years was an ethnohistorically ignorant assumption. Can't you see bad scholarship here? 35 years is pretty solid, at least for modern times, and maybe applicable to ancient times as well. Otherwise the standard 25 years seems correct. Seielstad used both times in his calculations and for my graph I used the dates from the 25 assumption, not the 35. I haven't done anything underhanded, despite Stewart's repeated accusations.

I didn't have to conveniently leave out "within modern times." That phrase does not change my statement. This is another of Stewart's rhetorical bamboozles.

[Tarski]

If Stewart had gone and read the Tremblay paper, which Seielstad cited, then Stewart wouldn't have made the ridiculous assertion that 25 or 35 years was an ethnohistorically ignorant assumption. Can't you see bad scholarship here? 35 years is pretty solid, at least for modern times, and maybe applicable to ancient times as well. Otherwise the standard 25 years seems correct. Seielstad used both times in his calculations and for my graph I used the dates from the 25 assumption, not the 35. I haven't done anything underhanded, despite Stewart's repeated accusations.

I didn't have to conveniently leave out "within modern times." That phrase does not change my statement. This is another of Stewart's rhetorical bamboozles.

Exactly. And, as I point out above, leaving out the "at least" in the phrase "at least in modern times" changes the sense. The actually sense seems to me to be that our confidence is higher for modern times but it is still thought to be true further in the past. Why leave out the "at least"?

Why don't those like Juliann find something nefarious in that? Bias?

[juliann]

What's really touching here is the way you guys say nothing, nothing, about my posts, but you all jump for joy when your champion makes a post. I'm sure you know what you are talking about CI, and of course you are a totally neutral judge.

Just out of curiosity and to supply the context...how many of your supporters have said anything neutral let alone positive about Stewart on your other boards? Nincompoop probably looks pretty good in comparison to what they say...which may be why you are bringing that kind of discourse over here. I pretty sure I can say that you would not find one of "us guys" who would not tell you that your knowledge is impressive and that you have created a good debate on many points. I have no doubt that you have made Stewart refine his position....as he has yours.

But again...if you are going to complain about your treatment here, don't leave out the rest of the story.