Quote :

"^ Now that we've gone down that fucking rabbit hole, do you believe that our planet is still in peril as described by Gore, the IPCC, and others or not?"

It's worth pointing out that it's not a "rabbit hole". How do you expect to make sense of these findings if you don't understand basic statistics?

10/12/2009 9:09:23 PM

i mean, you're the one that redefined "correlation in and of itself" to be "correlation plus extra information" in order to appear to win an argument

10/12/2009 9:10:45 PM

Quote :

"A lot of times the covariance matrix you're working with gives you causal structure. Not saying that any of these studies are in such a position (seeing as how I haven't read them), but it's certainly true of a lot of studies. In some cases where there's ambiguity about the causal direction in a correlation, you can orient the relation with prior knowledge (or temporal knowledge; A occurs before B). In some cases, there's no ambiguity about the direction of the causal relation implied by the correlation due to the other correlations/covariances present in the data."

^ Read it again and tell me I did that with a straight face

10/12/2009 9:13:43 PM

Oh, I'm sorry... I was talking about this:

Quote :

"sometimes correlation => causation there's your free science lesson for the day you can thank me later"

I believe the quote you referred to was a well executed backtrack:

Quote :

"A lot of times the covariance matrix you're working with gives you causal structure. Not saying that any of these studies are in such a position (seeing as how I haven't read them), but it's certainly true of a lot of studies. In some cases where there's ambiguity about the causal direction in a correlation, you can orient the relation with prior knowledge (or temporal knowledge; A occurs before B). In some cases, there's no ambiguity about the direction of the causal relation implied by the correlation due to the other correlations/covariances present in the data."

10/12/2009 9:16:01 PM

I said SOMETIMES correlation => causation.

What about that is a "well-executed backtrack"? I cited an entire FIELD of research in statistics, which is what I was referring to when I said "sometimes" above. There are, in fact, cases where the correlative structure can recover causal structure (it has to do with independence tests). What about this is unclear from anything I said? Do you honestly think I made an error and then went out and found a field of stats to cite such as to backpedal?

10/12/2009 9:18:36 PM

If you meant correlative structure => causal structure, then perhaps that is what you should have written.

10/12/2009 9:20:04 PM

...

I'm not understanding the distinction you're trying to draw. What would you call "correlative structure" other than the correlation matrix implied by the data? What would you call "causal structure" other than the causal graphs implied by the data?

10/12/2009 9:22:20 PM

in fact, correlation != correlative structure and causation != causal structure.

10/12/2009 9:23:16 PM

What distinction would you draw?

10/12/2009 9:27:28 PM

Dammit man, go start a mathematics thread!

Quote :

" A possible cooling trend is not new: http://www.newscientist.com/blog/environment/2008/05/stay-cool-about-short-term-climate.html As is mentioned in both articles, they believe it is temporary and warming will resume within a decade or so."

O Rly? Cooling trend not new?! Of course not. There was the Little Ice Age, the warming period that peaked in the 30s, the cooling trend that followed that, and now the warming period that peaked in the 90s (and ended in 1999).



So basically what they're saying (and what you're believing) is that a cooling period that last 20 years or more is nothing to be worried about but that a warming period that only lasted about 20 years is?

Yeah that makes perfect sense.

[Edited on October 12, 2009 at 9:30 PM. Reason : Levitus, S., G. Matishov, D. Seidov, and I. Smolyar (2009), Barents Sea multidecadal variability, Ge]

[Edited on October 12, 2009 at 9:32 PM. Reason : k]

10/12/2009 9:27:52 PM

@solinari This is the problem talking about this shit with people who are clueless and used to arguing like fucking pedants on a message board. I get to have semantic arguments about "correlative structure" vs. "correlation" with a couple of fucking goons who don't know what it means either way. I'm referring to a well-defined mathematical object (a correlation matrix). I can't imagine using these two terms to refer to different things, seeing as how I (and practically everybody else I know who does research in a similar area) uses these terms interchangeably. Perhaps you could point out to me what the difference is, what you had in mind, and how what I said is confusing under that interpretation? Terminology is not exactly standardized in this field, so I'm willing to give you the benefit of the doubt if you come from a fringe field; still, I'm unsure how you could have missed a charitable reading of what I said (under the assumption you have a clue what you're talking about).

[Edited on October 12, 2009 at 9:33 PM. Reason : .]

10/12/2009 9:32:57 PM

Quote :

"So basically what they're saying (and what you're believing) is that a cooling period that last 20 years or more is nothing to be worried about but that a warming period that only lasted about 20 years is?"

No, not at all.

1. Who's "they"? Different sources have varying predictions.
2. You don't know what I believe
3. No one has given an opinion about whether a particular cooling period would be "nothing to worry about."

[Edited on October 12, 2009 at 10:33 PM. Reason : .]

10/12/2009 10:32:12 PM

Yes, I know it is Fox. Lets get that out of the way right off the bat.



Anyway, thoughts? Comments?

10/12/2009 10:38:44 PM

^nothing new there, that kind of censorship behavior is pretty standard fare.

^^given that its been happening continually since the birth of the planet I don't really know why anyone would worry. Unless it's the onset of another ice age.

10/12/2009 11:16:06 PM

I'd say the biggest player in all this is being neglected by you in your historical observation--

automobiles.

10/12/2009 11:19:48 PM

^^That is one of the stupidest things I've ever read someone say about climate change.

I'm going to quote this:

Quote :

"given that its been happening continually since the birth of the planet I don't really know why anyone would worry. Unless it's the onset of another ice age."

[Edited on October 12, 2009 at 11:24 PM. Reason : seriously, that's like Youtube comments level of stupid]

10/12/2009 11:22:05 PM

Quote :

"I'm unsure how you could have missed a charitable reading of what I said "

Since when is someone obligated to give his opponent the benefit of the doubt in a political debate? You were spinning and got called out. That's all. Its no big deal... You've since clarified yourself and qualified your original statement to mean much less than it originally did on its face. That's how spin works, after all. State one broad powerful quote for maximum impact and then come back in with clarifying nuance later on after the emotional impact has run its course. (an apropos emoticon if there ever was one)

[Edited on October 12, 2009 at 11:24 PM. Reason : s]

10/12/2009 11:23:04 PM

^ Are you seriously this fucking stupid? First of all, it was a mathematical discussion (not a political debate). Second of all, it was clear I was talking about a restricted set of cases. Correlation != Causation is a tired ass cliche that people repeat when they don't know a fucking thing about statistics.

BTW: Still waiting for you to answer my questions. What meaning did you assign to "correlative structure" that had you in knots over what I said?

[Edited on October 12, 2009 at 11:55 PM. Reason : .]

10/12/2009 11:45:42 PM

Quote :

"Noel Keenlyside of the Leibniz Institute of Marine Sciences in Kiel, Germany, and colleagues used sea surface temperatures recorded over the last 50 years to constrain other variables in an existing climate model. They validated the approach by "hindcasting", that is, by checking that the model predicted the right conditions over recent history. Then to the forecasting: they reckon that over much of Europe and North America, the coming decade will see cooling, while Pacific regions remain the same temperature."

Quote :

"And it is, after all, a model ? built on assumptions and approximations that are not so different from the ones made by your local weather forecaster"

Quote :

"But the big picture, both in terms of time and of policy, shouldn't be forsaken in light of this educated guess at a close-up."

i like how this blogger denounces the methods used to formulate the cooling models, when they're the same methods used for all the AGW predictions

10/13/2009 12:43:27 AM

Quote :

"It's worth pointing out that it's not a 'rabbit hole'. How do you expect to make sense of these findings if you don't understand basic statistics?"

McDouche

We had a perfectly civil exchange, but you just had to come back and make you point shitty, didn't you? You smug asshole.

I will now deconstruct your pitiful attempt at a post. First, concerning "rabbit hole," it was word play--that you apparently aren't smart enough to get.

rabbit hole:

Quote :

"This usage has helped make the phrase refer more generally to any portal into a different or strange world. . . ."

http://en.wikipedia.org/wiki/Rabbit_hole

Indeed--the "different" and "strange" world of McDoucheland.

Second, you presuppose that I "don't understand basic statistics." Yet, (1) you have offered no proof of this and (2) I have, in fact, successfully completed college-level statistics. So, I have presented more evidence that I understand basic statistics than you have that I do not.

Third, you have not proved that an understanding of basic statistics or beyond is even necessary to grasp the "findings" at issue. As just about everyone knows, many of these reports are written in plain English.

Fourth, periods should be placed inside closing quotation marks--not outside. Are you from England?

Now, having thoroughly owned you--as I always do--can we get back on the fucking topic?

[Edited on October 13, 2009 at 5:41 AM. Reason : Thanks. ]

10/13/2009 5:39:00 AM

Quote :

"Second of all, it was clear I was talking about a restricted set of cases. "

Actually it was clear that you said, "correlation => causation" thereby maximizing the psychological impact of your point. You then came back afterwards and qualified your statement. This is how someone performs "spin"

Quote :

"Correlation != Causation is a tired ass cliche that people repeat when they don't know a fucking thing about statistics."

I agree that it is a tired ass cliche.... Perhaps because so many people tirelessly attempt to prove their points using simple correlation (re. drug restrictions, healthcare, global warming, taxation, etc. etc.)

You may be surprised to re-read the last two pages and note that no one has actually disagreed with your amended point that correlative structure => causal structure.

As I said, this discussion boils down to the fact that you used "correlation" as shorthand for "correlative structure" in an attempt to spindoctor the discussion. You got called out on it, got pissy, and now you won't admit that correlation != correlative structure, which makes you look like a slobbering fucking dunce.

[Edited on October 13, 2009 at 8:49 AM. Reason : s]

10/13/2009 8:24:27 AM

Quote :

"Fourth, periods should be placed inside closing quotation marks--not outside. Are you from England?"

actually, this isn't always true...when the quoted text is a complete statement, yes, the period should be inside...when the quoted text is a fragment (be it a phrase or a single word or whatever), the period should NOT go inside (though it is still acceptable if it does)...correct examples:

"Johnny planted apple trees."

Johnny planted "apple trees".

10/13/2009 8:30:56 AM

^ Wrong. Wow.

Quote :

"The MLA rule is that periods and commas ALWAYS go inside the closing quotation marks. Semi-colon and colon always go outside. The question mark and exclamation mark go where appropriate to fit the sense of the original (if the question is part of the quotation, it goes inside). Students usually follow the question mark rule (following whether it is there in the original) for period and comma, and you should not."

http://www.jsu.edu/depart/english/gates/faqcites.htm

Quote :

"Periods and commas, whether or not they are part of the quoted material, always go inside the closing quotation marks. . . ."

http://www.cws.illinois.edu/workshop/writers/tips/quotations/

Now, can we please get back on the fucking topic?

10/13/2009 8:50:35 AM

Quote :

"I'd say the biggest player in all this is being neglected by you in your historical observation-- automobiles."

WillemJoel, I don't really see the point in this statement. Biggest player? Taken together as one source, automobiles are not the biggest emitter of CO2. And that's besides the point, as there is no direct evidence that CO2 caused the warming from 1980-1999.

carzak, don't be ridiculous. Please tell me why natural climate shifts over history are now the fault of humans, that we should worry about them, and that "we" should do something (as if we even could).

10/13/2009 9:15:52 AM

Quote :

"Wrong. Wow."

i said it "isn't always true"...you really should admit that the brit standard makes more sense than the american (after all, adding something to a quotation that wasn't actually IN the ORIGINAL quote defeats the purpose of quoting in the first place, don't you think?)

in any case, in an example like this, even by MLA standards, the period goes outside:

This semester, I received four "As" and a "B".

so, really, your use of "always" is incorrect while my use of "always" is

[Edited on October 13, 2009 at 9:24 AM. Reason : .]

10/13/2009 9:24:35 AM

^ You're just flat wrong. And according to MLA, letters that are referred to as letters should be underscored (or italicized).

And I won't discuss this issue further. You're just trolling.

[Edited on October 13, 2009 at 9:40 AM. Reason : V It's on p. 95, ch. 3.3.1.]

10/13/2009 9:31:53 AM

Quote :

"You're just flat wrong. And according to MLA, letters that are referred to as letters should be underscored (or italicized)."

link? or opinion?

Quote :

"And I won't discuss this issue further. You're just trolling."

TSB at its finest!

carry on.

10/13/2009 9:38:07 AM

Quote :

"Underline (or italicize) 'words and letters that are referred to as words or letters' (Gibaldi 95). For example, write 'The term American Indian is inclusive of over 500 Federally recognized ethnic communities.'"

http://www.docstyles.com/mlacrib.htm

10/13/2009 9:43:43 AM

Quote :

"Actually it was clear that you said, "correlation => causation" thereby maximizing the psychological impact of your point. You then came back afterwards and qualified your statement. This is how someone performs "spin""

Quote :

"nope but sometimes correlation => causation"

...

You're really reaching and you look fucking foolish. Hey by the way tell me what you think "correlative structure" means so that I can figure out why you didn't understand what I said (hint: I know why you didn't understand it: it's because you haven't a fucking clue what you're sputtering on about).

Quote :

"Indeed--the "different" and "strange" world of McDoucheland. Second, you presuppose that I "don't understand basic statistics." Yet, (1) you have offered no proof of this and (2) I have, in fact, successfully completed college-level statistics. So, I have presented more evidence that I understand basic statistics than you have that I do not. Third, you have not proved that an understanding of basic statistics or beyond is even necessary to grasp the "findings" at issue. As just about everyone knows, many of these reports are written in plain English. "

I like how you flew off the handle when I wasn't even talking about you in particular. I'm saying that the diversion wasn't a trip down the rabbit hole at all -- I was merely pointing out a side issue, and the idiot crew (Solinari, aaronburro) decided to try and "pwnzor" me, revealing that they know next to nothing about stats. Correcting them was important because how the hell are you supposed to critically read a paper that SOLELY consists of statistical analysis without even an understanding of the relationship between correlation and covariance? This is what I was pointing out, but your fragile ego caused you to spin off into a tirade.

Quote :

"Fourth, periods should be placed inside closing quotation marks--not outside. Are you from England? "

No but I do mathematical and scientific writing and that's how it's typically done. Not that you'd know.

Quote :

"You may be surprised to re-read the last two pages and note that no one has actually disagreed with your amended point that correlative structure => causal structure. As I said, this discussion boils down to the fact that you used "correlation" as shorthand for "correlative structure" in an attempt to spindoctor the discussion. You got called out on it, got pissy, and now you won't admit that correlation != correlative structure, which makes you look like a slobbering fucking dunce."

What does "correlative structure" refer to, in your opinion? I was talking about the matrix of correlations implied by the data, which was obvious to anybody here with a post-high school education in mathematics. There is no real distinction between saying "sometimes correlation => causation" and saying "sometimes correlative structure => causal structure". It's not "short hand" and "spin doctoring" -- it's what the fucking words refer to. I'm saying in many cases you can take a correlation matrix and with the conditional independence relations implied by it, recover a graphical structure with a causal interpretation (causation; causal structure). Of course I only happen to know what I'm talking about here, so feel free to ignore me.

Again: tell me what "correlative structure" refers to, in your opinion. That'll straighten this right up. Or: you can continue to ignore me and put it off since you don't have a fucking clue. This is what happens when message board idiots try to argue mathematics by using typical slippery semantic tricks and pedantry. People who know what the fuck is up are not fooled by your stupid shit.

Quote :

"^ You're just flat wrong. And according to MLA, letters that are referred to as letters should be underscored (or italicized). And I won't discuss this issue further. You're just trolling."

It really depends on the field to be honest. In fields that make a hard and fast distinction between use and mention (like philosophy, mathematics, linguistics, etc), people tend to drop the period after the quotation mark (unless, of course, they're quoting a full sentence). The reason why is because they don't want to include the period in the linguistic object to which they're referring.

[Edited on October 13, 2009 at 11:25 AM. Reason : .]

10/13/2009 11:19:59 AM

Quote :

"There is no real distinction between saying "sometimes correlation => causation" and saying "sometimes correlative structure => causal structure"."

Well, actually there is.

10/13/2009 12:51:37 PM

Until you tell me what the difference is, you're bullshitting. Have fun with your private language though.

[Edited on October 13, 2009 at 12:56 PM. Reason : .]

10/13/2009 12:53:55 PM

I guess by private language, you mean the english language... Sure, I'm having fun with it, but apparently not as much fun as you.

10/13/2009 1:19:15 PM

Define "correlative structure".

10/13/2009 1:25:44 PM

correlative structure includes more information than the simple measure of correlation between two variables.

do you deny this?

[Edited on October 13, 2009 at 1:39 PM. Reason : s]

10/13/2009 1:38:58 PM

Quote :

"correlative structure includes more information than the simple measure of correlation between two variables. do you deny this?"

It's the measure of correlations between all of the variables you're measuring, whether 1 or arbitrarily many.

Why not just give me a definition you fucking moron? This isn't Socrates-hour, this is you fessing up.

[Edited on October 13, 2009 at 1:42 PM. Reason : .]

10/13/2009 1:42:08 PM

no reason to get all pissy because I called you out.

10/13/2009 2:05:44 PM

Well at least I can reasonably tell you're trolling at this point. Either way you're a shit head and hilariously wrong.

10/13/2009 2:07:01 PM

no, you're the troll, for claiming that a correlative structure for one variable can allow the inferrence of causality

10/13/2009 2:17:53 PM

Never once did I claim or intend to claim that. You could force that interpretation out of my words, I suppose, but rarely do people associate "correlation" with "self-correlation" as you're trying to do here. If you only measure one variable, you'll have a 1x1 correlation matrix with the entry "1" in it.

You may be able to make inferences about the inputs in such a case, but you won't run a statistical study if you're only measuring a single variable (usually). Such cases are controlled laboratory experiments where you're measuring the effects of certain actions on certain controlled outcomes (experiments with incline planes and shit).

10/13/2009 2:21:03 PM

Quote :

"correlation => causation"

Quote :

"There are, in fact, cases where the correlative structure can recover causal structure"

Quote :

"It's the measure of correlations between all of the variables you're measuring, whether 1 or arbitrarily many."

Quote :

"Never once did I claim or intend to claim that."

Look man... I'm not saying this is what you really claimed. I've said all along that you were just spinning in the well worn manner of making a very broad assertion with sweeping language for maximum impact followed by qualifications after the fact. You're still crying about it, though... Not sure why. I guess its the first time someone called you out for gross exaggeration?

10/13/2009 2:52:27 PM

Is there any particular reason you quote only part of what I said? Oh that's right because you're wrong and the only way you can hold together your fragile internet image is to win debates by mis-quoting your opponent. Lols.

You're an ECE guy? Fuck, I hope you get your head together and learn some math.

10/13/2009 2:58:39 PM

I'm not misquoting you... Here's why you had to employ these spin tactics.... If you hadn't, the conversation would've gone like this:

Person: <quotes article>
Me: correlation != causation
You: correlation matrix sometimes allows someone to infer causal relationships
Me: Does the study in question use a correlation matrix
You: Not saying that any of these studies are in such a position (seeing as how I haven't read them)
Me: Then STFU
You: dangit... should've used spin tactics instead.

[Edited on October 13, 2009 at 3:02 PM. Reason : s]

10/13/2009 3:01:23 PM

[unsubscribe]

10/13/2009 3:02:54 PM

So now that I've determined you're trolling it doesn't bother me so much anymore.

But man, if you're actually serious about anything you're saying? I hope you find a comfortable office job where you're not expected to think.

Let's get back to the discussion of statistical data, except let's ensure that the only people discussing are people that don't understand basic statistics!!!

[Edited on October 13, 2009 at 3:03 PM. Reason : .]

10/13/2009 3:03:00 PM

well i wrote you off as a troll about the time you said, "correlation => causation" so nyah!!

Quote :

"Let's get back to the discussion of statistical data"

There never was one.

[Edited on October 13, 2009 at 3:04 PM. Reason : s]

10/13/2009 3:03:55 PM

Because analyzing global weather trends doesn't require any knowledge of statistical inference

You know what, it's clear you're trolling. I'm done. It's too bad aaronburro has conditioned me to expect drooling dunces in this section.

[Edited on October 13, 2009 at 3:10 PM. Reason : .]

10/13/2009 3:08:12 PM

After two pages of denial, you still won't admit that you were trying to spin the issue and got caught. I believe that's the textbook definition of a troll. GG - you got me.

10/13/2009 4:57:02 PM

Quote :

"As is mentioned in both articles, they believe it is temporary and warming will resume within a decade or so."

So, it's temporary cooling that is so powerful that it is counteracting the massive effect that CO2 should be having? really?

10/13/2009 7:29:07 PM

^in one of the blogs "articles" the blogger is critical of the temporary cooling predictions because they're based on models and predictions which aren't concrete enough for him...except when the models and predictions predict AGW

10/13/2009 7:40:30 PM

its only credible if you agree with it

10/13/2009 8:38:47 PM