How many is a "Consensus"?

All I know is all this global warming is murder on my heating bill. Brrrr.

Successfully updated karma reason!

Q: How many is a scientific consensus?

A: One politician.

scientific consensus is not achieved by scientists... it is achieved by political activists with an axe to grind.

Amusingly... it took the author of the "study" throwing out 97% of the results (well 97.43%) to achieve a 97% consensus.

Successfully updated karma reason!

Quoting Nitro Cruiser,

reply 1

All I know is all this global warming is murder on my heating bill. Brrrr.

Heating and cooling. The new religion says whatever the weather today (hot or cold), Global Warming is to blame. Kidn of like the religious saying "the devil made me do it". AGW made Gaia do it!

Successfully updated karma reason!

Quoting taltamir,

reply 2

Q: How many is a scientific consensus?

A: One politician.

scientific consensus is not achieved by scientists... it is achieved by political activists with an axe to grind.

Amusingly... it took the author of the "study" throwing out 97% of the results (well 97.43%) to achieve a 97% consensus.

Do not tell the religious of "consensus". To them, nothing else matters. All data that does not fit - must be shit.

Successfully updated karma reason!

Speaking of that, there's this.

Successfully updated karma reason!

Quoting Daiwa,

reply 5

Speaking of that, there's this.

Not surprising. Some people still need to look at themselves in the mirror. If you want to read a Warmist that is honest, read Judith Curry (Climate Etc.). She is getting heat from the rabid wing of the religion, but while she does believe in AGW, she is not a kool aid drinker and does listen to both sides.

Successfully updated karma reason!

Quoting ,

quoting post

Now comes news of a study done on the survey. It appears that while over 10,000 scientists were asked to complete the survey (only 2 questions) only about 30% did. Still not a bad number, 3000, right? Wrong. The "researchers" then started whittling the number down since the results were not to their liking. So after removing every group imaginable they managed to find a group that consisted of 97% that agreed that man had "some" impact on the climate (note not even a major part of it).

I'm guessing none of you did even bother to click on the link provided in the article and read the paper. If you had you would have seen there's nothing to be outraged about. From the actual paper:

Results show that overall, 90% of participants answered “risen” to question 1 and 82% answered yes to question 2. In general, as the level of active research and specialization in climate science increases, so does agreement with the two primary questions (Figure 1). In our survey, the most specialized and knowledgeable respondents (with regard to climate change) are those who listed climate science as their area of expertise and who also have published more than 50% of their recent peer-reviewed papers on the subject of climate change (79 individuals in total). Of these specialists, 96.2% (76 of 79) answered “risen” to question 1 and 97.4% (75 of 77) answered yes to question 2. This is in contrast to results of a recent Gallup poll (see http://www.gallup.com/poll/1615/Environment.aspx) that suggests that only 58% of the general public would answer yes to our question 2. The two areas of expertise in the survey with the smallest percentage of participants answering yes to question 2 were economic geology with 47% (48 of 103) and meteorology with 64% (23 of 36).

So they didn't cherry-pick their favourite responses according to some weird unrelated criteria just so they'd get the response they wanted. They picked the responders who had "climate science" as their speciality and who had recently published peer-reviewed studies on the subject of climate change. Not to mention that 82% of all the responders answered that humanity has a significant impact. Seeing as most would view that as a consensus, no doctoring was necessary to get one. They weren't out to invent some kind of following for global warming, they were out to find how large said following was among climate change experts. (Granted 77 is still a low figure, but if only about 2% of earth scientists are climatologist that group would be quite small to begin with.)

Successfully updated karma reason!

Quoting Wieke,

reply 7

So they didn't cherry-pick their favourite responses according to some weird unrelated criteria just so they'd get the response they wanted.

Cherry picking the answer based upon some predetermined bias, or some "weird unrrelated criteria" is stil lcherry picking. In other words, the consensus is bogus as is the whole claim. It does not matter if they attribute their actions to divine intervention, of that they hate strawberries.

true polls are based on scientific principals of statistics - not 'arrive at conclusion, then validate it" type of hokum.

Successfully updated karma reason!

So you're cherry picking if you set out to "assess the scientific consensus on climate change through an unbiased survey of a large and broad group of Earth scientists", then break down your data into several categories (based on expertise regarding climate change) and report the stats of all categories?

There was no obfuscation or intentional misleading in the paper. They reported every manipulation of the data and it's consequences. Not to mention it was an peer-reviewed study published by the American Geophysical Union. It wouldn't have gotten that far if it was just a case of "arrive at conclusion, then validate it".

Successfully updated karma reason!

Quoting Wieke,

reply 9

So you're cherry picking if you set out to "assess the scientific consensus on climate change through an unbiased survey of a large and broad group of Earth scientists",

Your premise is flawed. They did not set out to do anything with an "unbiased" survey as the wording on the survey was done to elicit the exact response they reported. That is your first error. Second error - regardless of your intent, if your methodology is flawed, then it is not unbiased or meaningful. A true survey would have targeted the sample group and reported the results of said group. They used a shotgun blast and an incorrect data gathering methodology.

Quoting Wieke,

reply 9

There was no obfuscation or intentional misleading in the paper.

There was gross obfuscation and intentional misleading. But then I never said it was in the paper. The paper was an exercise in sophistry. However the use of the paper is where the gross obfuscation and misleading comes in. The paper is merely a peacock preening, and has no scientific value whatsoever. However, whether intentional or not on the part of the authors, the paper has been used as an authoritative source for the "97%" figure ever since. A figure that is has no basis in fact (other than to say 97% of the winnowed group of the non-scientific biased poll).

Quoting Wieke,

reply 9

Not to mention it was an peer-reviewed study published by the American Geophysical Union.

You can peer review a grocery list - that does not make it science. The paper was not peer reviewed for statistical accuracy as there is no statistics involved in it! And no statistician worth a lick of salt would ever sign his name to the paper as being more than a beauty contest.

Perhaps you should read up on what statistics is (it is a real science, not what ERA the Pitcher on your Baseball team has). Clearly you do not know what or how to go about either calculating or obtaining true statistics.

Successfully updated karma reason!

Quoting Dr Guy,
reply 10

Perhaps you should read up on what statistics is (it is a real science, not what ERA the Pitcher on your Baseball team has). Clearly you do not know what or how to go about either calculating or obtaining true statistics.

Well couple of months I finished a course on statistics and I must have been unconsciously suppressing it, because you're right. There isn't a single trace of any real statistical analysis to be found in that paper (no p value, correlation, analysis used etc). Then again they didn't set out to analyse the effect/correlation/relation between two or more variables, just how high a single one was. So I'm unsure how much statistical analyse can be done using only a single nominal variable. (I guess they should have determined the effect of specialization and publishing activity on the answer on both questions, at least then they would've known if the difference between subgroups was significant.)

Quoting Dr Guy,

reply 10

Your premise is flawed. They did not set out to do anything with an "unbiased" survey as the wording on the survey was done to elicit the exact response they reported. That is your first error. Second error - regardless of your intent, if your methodology is flawed, then it is not unbiased or meaningful. A true survey would have targeted the sample group and reported the results of said group. They used a shotgun blast and an incorrect data gathering

Quoting that paper,

The objective of our study presented here is to assess the scientific consensus on climate change through an unbiased survey of a large and broad group of Earth scientists.

Well that was a quote from the paper, so even if that wasn't what they did (either through malice or incompetence), it is what they reported to be what they set out to do.

The only kind of sample bias I can perceive comes from the fact that they got their participants from a Directory of Geoscience Departments (the 2007 edition of this one) and that participation was voluntary (would be interesting to see if there is a significant interaction between the "could/couldn't be bothered" and "answer to the question" variables, but you can't force people to participate). And unless there is a massive population of unlisted maverick climatologist out there, it wouldn't have had much influence. Granted they didn't report if the participants were picked at random and seeing as the AGU lists at least 120,000 earth scientist as members this could be a serious source of bias (though I'd give them the benefit of the doubt an assume they picked at random).

I'm not sure I'm getting the point of your last 2 sentences. Are you claiming that if you have data on a large group you are only allowed to report on the group as a whole? That you shouldn't report on subgroups that are completely contained within the larger group (even if you are also reporting on the group as a whole)?

Granted you and the financial post were right about the questions. Those were worded poorly.

There was gross obfuscation and intentional misleading. But then I never said it was in the paper. The paper was an exercise in sophistry. However the use of the paper is where the gross obfuscation and misleading comes in. The paper is merely a peacock preening, and has no scientific value whatsoever. However, whether intentional or not on the part of the authors, the paper has been used as an authoritative source for the "97%" figure ever since. A figure that is has no basis in fact (other than to say 97% of the winnowed group of the non-scientific biased poll).

Is it really that misleading to report that “97% of the world’s climate scientists”, only failing to mention these only include the opinions of actively publishing climate scientists with 50% of their recent publications covering climate change?

You can peer review a grocery list - that does not make it science. The paper was not peer reviewed for statistical accuracy as there is no statistics involved in it! And no statistician worth a lick of salt would ever sign his name to the paper as being more than a beauty contest.

Well I'm hoping a peer-review by a scientific journal/publisher does mean something these days (depending on the reputation of said journal/published). But I have yet to venture into the world of academia and may be a bit naive (I'm just a student).

EDIT: I know it's off-topic but what's the exact quote syntax? I don't seem able to get the last 2 quotes in proper quote blocks without it somehow getting included in the first quote block (or removing the "who = ..." bit).

Successfully updated karma reason!

Dr. Guy's being a bit harsh, but I like this discussion. And I agree that it's what's been done with the bloody shirt that's the issue, more than the bloody shirt itself.

Still, 97% of a sample size of 79 out of 10,000 is pretty much meaningless. The American Geophysical Union can probably adequately peer-review papers on geophysical science but I doubt they have any expertise in polling, which is all the paper was - an opinion poll. I'm a little surprised it was accepted for publication unless it was clearly identified as a survey piece.

For anyone to take that paper and use it as the basis for a conclusion that "97% of the world's climate scientists..." believe anything is just dishonest. The honest conclusion is that 75-76 self-identified climate scientists agreed with each other on the answers to two questions.

As for the quote syntax, it's a little tough to 'show' it without it being invoked so I'll use the word 'here' to substitute for the word 'quote' - it's a leading bracketed tag - [here] and a closing bracketed tag with a preceding forward slash [/here] enclosing the text you wish to quote.

Successfully updated karma reason!

Before the numbers, it might be easier to solve the issue here ... When to Doubt a Scientific ‘Consensus’.

http://www.american.com/archive/2010/march/when-to-doubt-a-scientific-consensus/

It is of course one sided ... but the steps themselves apply to all.

Successfully updated karma reason!

Quoting Wieke,

reply 11

The only kind of sample bias I can perceive comes from the fact that they got their participants from a Directory of Geoscience Departments (the 2007 edition of this one) and that participation was voluntary (would be interesting to see if there is a significant interaction between the "could/couldn't be bothered" and "answer to the question" variables, but you can't force people to participate).

Woo Boy! You blew a fat one Gordie! Their sample bias was in the their method of collection! It was not random, and the respondents were not random. So they blew it out of the gate.

At the very least, they should have reported all the no-shows as "no opinion" but that would not have made it valid - just more accurate), so that 97% figure would be about 3%, with a very large "Don't Know' percentage what would not have made it worth reporting.

Quoting Wieke,

reply 11

I'm not sure I'm getting the point of your last 2 sentences. Are you claiming that if you have data on a large group you are only allowed to report on the group as a whole?

NO, you are allowed to report on the subset - with qualifications. From the get go, I have not slammed the authors of the preening page for anything more than running a fixed game. My major slam was on the people who are using the number which is virtually worthless (other than to relate the opinions of 75 totally not random people - and 77 is not a large enough sample size period).

Quoting Daiwa,

reply 12

Is it really that misleading to report that “97% of the world’s climate scientists”,

Yes

Quoting Wieke,

reply 11

Well I'm hoping a peer-review by a scientific journal/publisher does mean something these days

Based upon Climategate emails, I would say not much in this area. After all, we have the Jesus Paper - try explaining that one.

Successfully updated karma reason!

Quoting Wieke,

reply 11

EDIT: I know it's off-topic but what's the exact quote syntax? I don't seem able to get the last 2 quotes in proper quote blocks without it somehow getting included in the first quote block (or removing the "who = ..." bit).

I have found (I do not know if it is listed anywhere) that there are 2 "quirks" on quoting. One is that they all have to be the same type (all with who or all without). The other is there is a limit - but it is large - I think 15. Nesting is about 5 I think.

Successfully updated karma reason!

Quoting Daiwa,

reply 12

For anyone to take that paper and use it as the basis for a conclusion that "97% of the world's climate scientists..." believe anything is just dishonest. The honest conclusion is that 75-76 self-identified climate scientists agreed with each other on the answers to two questions.

The problem with reporting that way (from a human standpoint, not a science one) is that people normally will ask - "how does that relate to all"? And there's the sticking point - you cannot get to there from here. Just as those polls on CNN and Fox are basically worthless, so this one is as well for determining anything meaningful.

Successfully updated karma reason!

How many is a "Consensus"?

Devil is in the Details