and

http://noconsensus.wordpress.com/2011/03/04/kgnd-cross-validation-part-ii-parting-thoughts/

Pretty conclusive in my opinion.

Perhaps you could try escaping into the quibble that you can’t technically alter a fact that doesn’t exist. You never know, it might work. (No it won’t). It’s worth a try. (No it isn’t).

You still insist on avoiding any reference to the claim you are supposed to be defending, that Ryan O’Donnell, being of sound mind and body, and acting of his own free will, at 10:50 AM on or about Feb 11 2011, did assert claim and hold to be true that Steig (hereinafter known as Steig)in his second review, did specifically propose that O’Donnell (hereinafter known as Thingummybob)should use iridge. I can’t write it any plainer than that. The most cursory examination of this claim shows this claim to be:
a) An error
b) Disingenuous
c) A lie
I have no evidence or inclination to suspect it is anything other than an error. The lie came later.

Here is all the context you need. If O’Donnell doesn’t equate ‘iridge’ with ‘the best method’, then why does he accuse Steig of insisting that it be used, when Steig only ‘insisted’ on the best method? HE clearly conflates them as synonymous, hence his accusation, while Steig clearly DOESN’T, hence what he actualy wrote. ‘PERHAPS’ Steig doesn’t know on way or the other. Perhaps I can read. It was O’Donnell who did all the proposing. You can’t even assert that Steig did any seconding of that proposal. He clearly abstained.

Back of he net! One-nil to the Boro.

As near as damn it. But I’d put it a bit more succinctly. Something along the lines of “Two blokes are wrong about each other on the internet – hilarity ensues” would suffice.

I suspect that you have never been a reviewer (we call them referees in my discipline) for a journal. “Perhaps” is not a mere suggestion to make conversation. The job of a reviewer is to make recommendations for changes to a paper and to offer final advice to an editor who makes the decision on acceptance or rejection. Statements like “the results from the ‘iridge’ infilling should be used instead” (a statement which originated with Steig not with the authors) is a recommendation for the authors to make changes, pure and simple.

You somehow also have an off-the-wall idea that because the authors made an initial use of iridge for additional information, this somehow contradicted the possibility of Steig proposing that iridge become the central methodology used. You either can’t or you refuse to understand the context of the argument, so all you are capable of is sophistry which “wins conclusive points”, but only in your own mind. Get a new book.

(It’s like talking to a rock! What a waste of electrons…)

Ryan O’Donnell claimed that “Steig acting as Reviewer A, in his Second Review, had asked the editor to “insist” that we present the “most likely” West Antarctica trends, specifically proposing iridge” You have conceded that claim is not true, that in fact ‘the authors’ had already introduced iridge before Steig made any mention of it, so it can’t possibly have been his proposal.

Key words: “introduced” and “proposing.”

Key facts:
* The authors first mentioned the word “iridge”
* The authors used iridge as an additional test
* The authors proposed the possibility of using iridge in further future work, not in this paper.
* Steig’s insistent request (for “most likely” method) was interpreted both by the editor and the authors as an insistence that iridge be used for the main result. And his third response proves that he was in agreement with this.
* Steig also argued that iridge work should not wait for another paper but be used in this one.
* Steig admits he is not a statistician and thus is unable to weigh in professionally on the different methodologies used.

Two ways of looking at it:

Pro-Authors:
* Their use of iridge for confirmation did NOT “introduce” nor “propose” iridge as the primary method for the paper. Since all the methods produced similar results, this wasn’t even a big deal for them.
* Steig was the one who (insisted) on iridge for this paper, “softened” slightly by also using the word “perhaps.”
* Steig also argued that iridge analysis not be delayed to a future paper.
* It is disquieting that Steig would now claim iridge should not have been used, and particularly so since he also (separately) claims to not be able to make such pronouncements.

Pro-Steig:
* The authors first mentioned iridge, this “introduced” the topic. Doesn’t matter that they explicitly said full iridge analysis should wait for a future paper.
* Steig’s “insistence” and further recommendation that iridge work not be delayed was simply a desire for the best math to be used now rather than later.
* All the rest of this is obfuscation. Why pick on Steig?!

Do I have it about right?

Erm… nobody mentioned the underlying issue, and your crude attempt to shift attention away from your complete failure to support the claim being defended is duly noted. Let me remind you of it. Ryan O’Donnell claimed that “Steig acting as Reviewer A, in his Second Review, had asked the editor to “insist” that we present the “most likely” West Antarctica trends, specifically proposing iridge” You have conceded that claim is not true, that in fact ‘the authors’ had already introduced iridge before Steig made any mention of it, so it can’t possibly have been his proposal. Hey I’ve got an idea. Why not do a word search for ‘propose’ or ‘proposal’(don’t bother). You might find it (no you won’t).

“In his previous post on Steig’s Trick Ryan indicates the intent and focus of the paper:”
I never asked you to demonstrate the intent and focus of the paper. If there is anyone in need of intent and focus it is clearly you. Keep on topic. You took it upon yourself to defend a specific claim, and you are singularly failing to do so. Son I am dissapoint.

In reply, Steig made the statement that I quoted earlier:
“Perhaps, ,as the authors suggest, kgnd should not be used at all, but the results from the ‘iridge’ infilling should be used instead.

Erm… It was me who drew your attention to it, not you mine. It was the fist post you responded to. Keep up. Claiming my points as your own is hardly going to get you anywhere is it? I’m beginning to suspect you operate in an alternative reality. You’ll be claiming you showed me what a scoundrel O’Donnell is next.

I already explained this to you earlier. Steig made the incorrect statement that Ryan had “suggested” that the iridge approach become the main method in the paper.,But I wasn’t asking you to defend any incorrect statements by Steig. I was asking you to defend a specific incorrect statement by O’Donnell. Pay attention.

For the rest of the story from here, I “suggest” that you go read (and understand) the very clearly written post, by Steve on this very same issue.And I suggest you go back and read my posts, and address what I say, not what you wish I’d said. You seem to be labouring under the misaprehension that ‘Perhaps’ means
something other than ‘Perhaps’. In fact the whole tenor of your posts imply you think it means ‘definitely is’, whereas in my book it is exactly equivalent to ‘perhaps not’. And I think it only fair to tell you at this stage that my book is the Oxford English Dictionary, not the Boys-Own Book of Boils and Sores.

You can’t seem to grasp the underlying issue here. No one has claimed that the authors did not initially make use of ridge regression for some secondary purpose in responding to Eric’s first review. In his previous post on Steig’s Trick Ryan indicates the intent and focus of the paper:

The first salient point is that Eric still doesn’t get it. The whole purpose of our paper was to demonstrate that if you properly use the data that S09 used, then the answer changes in a significant fashion. This is different than claiming that this particular method (whereby satellite data and ground station data are used together in RegEM) provides a more accurate representation of the [unknown] truth than other methods. We have not (and will not) make such a claim. The only claim we make is – given the data and regression method used by S09 – that the answer is different when the method by which the data are combined is properly employed. Period.

Let me translate that for you. The purpose was to show that Steig used the TTLS in Regem method incorrectly. It was NOT to show that another method (e.g. ridge regression) works better. Period.

They purported to do that. Steig objected to things in the first review. Ryan et al. addressed the issues raised AND to add weight to their objections used Regem with the iridge option as additional evidence. As stated in the quote above, this was NOT intended to become the central focus of the paper.

In reply, Steig made the statement that I quoted earlier:

Perhaps, ,[b]as the authors suggest[/b], kgnd should not be used at all, but the results from the ‘iridge’ infilling should be used instead.

I already explained this to you earlier. Steig made the incorrect statement that Ryan had “suggested” that the iridge approach become the main method in the paper. That’s what “used instead” means. I indicated to you that Ryan did not make this suggestion anywhere in any document – it originated in Eric Steig’s mind. Got it?

For the rest of the story from here, I “suggest” that you go read (and understand) the very clearly written post, by Steve on this very same issue.

My bad. I thought ‘Testing by infilling ground stations using ridge regression’ meant ‘Testing by infilling ground stations using ridge regression’. But I’m not a climatologist.

So, according to you, it means ‘NOT Testing by infilling ground stations using ridge regression’ or perhaps ‘Testing by infilling ground stations by NOT using ridge regression’ does it? So did the authors tell some ikkle porkies about using it back then, or is someone telling ikkle porkies about not using it now? I’m, confused, at what stage are they telling porkies? Did they or did they not use ‘testing by infilling ground stations using ridge regression’ as they claim they did?

The concept of before and after seems to be escaping you. Look it up and have another go. If you’re having trouble with which order they go in I’d be more than happy to assist you. It’s a simple task facing you. The specific claim by O’Donnell is that Steig specifically proposed they use iridge in his second review, and the evidence posted by O’Donnell himself in support of that claim clearly shows Steig commenting of ‘the authors’ use of it already! LOL. Furthermore, the claim that Steig insisted it was the best method is hardly supported by him suggesting ‘Perhaps’ it wasn’t the best method. Whoops-a-daisy!

And then to cap it off, all that you can post in response to that refutation is proof positive that the authors had indeed already used iridge prior to any mention of it by Steig!!! Epic fail mate. I’m sorry to have to break this to you, but your evidence is supposed to support your case, not mine. That’s how it works, supposedly. But hey, have another go. You know you want to.

Perhaps, ,[b]as the authors suggest[/b], kgnd should not be used at all, but the results from the ‘iridge’ infilling should be used instead.

I searched the entire 47 page Response to Reviewer A, and I could not find any place where the authors suggest that “the results from the ‘iridge’ infilling should be used instead“.

In fact, in their response to Steig, they state:

Testing by infilling ground stations using ridge regression with the ridge parameter for each time step determined via generalized cross validation yields validation statistics, spatial patterns and West Antarctic trends (~0.11) comparable to our main results, and performing RLS reconstructions after offsetting stations based on periods of mutual overlap (i.e., no infilling) yields validation statistics, spatial patterns and West Antarctic trends (~0.10) also comparable to our main results (¶4.C – U, 4.Y – AE). These additional tests – which form the basis for a future work – have now been incorporated into the main text.

In plain English, it is quite clear that this is additional support to the case against the Steig paper. It also takes quite an imagination to turn that into your interpretation of the Steig quote that “introduction” of this material is tantamount to a suggestion by the authors that it become a replacement of the prior criticism of the Steig methodology rather than supplementary evidence.