Tol statistically deconstructs the 97% Consensus

Dr. Richard Tol has been tweeting a statistical destruction of the “97% consensus” study, Cook et al. (2013) by educating co-author Dana Nuccitelli as to why his “sample” is not representative.

In his defense, [Dana] has had limited exposure to stats at uni” – Richard Tol

Including “global” before “climate change”, Cook et al. dropped 75% of papers and changed disciplinary distribution.
Including “global” before “climate change”, Cook et al. dropped many papers by eminent climate researchers.
Including “global” before “climate change”, Cook et al. dropped 33 of the 50 most cited papers.
Choosing exclusive WoS over inclusive Scopus, Cook et al. dropped 35% of papers and changed disciplinary distribution.

As Dr. Tol so eloquently put it,

[Dana] I think your sampling strategy is a load of nonsense.” – Richard Tol

CV of Dr. Richard Tol:

M.Sc. Econometrics, Vrije Universiteit Amsterdam, Netherlands (1992); Ph.D. Economics (Thesis: “A decision-analytic treatise of the enhanced greenhouse effect“), Vrije Universiteit Amsterdam, Netherlands (1997); Researcher, Institute for Environmental Studies, Vrije Universiteit Amsterdam, Netherlands (1992-2008); Visiting Researcher, Canadian Centre for Climate Research, University of Victoria, Canada (1994); Visiting researcher, Centre for Social and Economic Research on the Global Environment, University College London, United Kingdom (1995); Acting Programme Manager Quantitative Environmental Economics, Institute for Environmental Studies, Vrije Universiteit Amsterdam, The Netherlands (1998-1999); Visiting Associate Professor, Department of Engineering and Public Policy, Carnegie Mellon University (1998-2000); Board Member, Centre for Marine and Climate Research, Hamburg University (2000-2006); Lead Author, IPCC (2001); Contributing Author and Expert Reviewer, IPCC (2001, 2007); Associate Editor, Environmental and Resource Economics Journal (2001-2006); Adjunct Professor, Department of Engineering and Public Policy, Carnegie Mellon University (2000-2008); Michael Otto Professor of Sustainability and Global Change, Department of Geosciences and Department of Economics, Hamburg University, Germany (2000-2006); Editor, Energy Economics Journal (2003-Present); Visiting Research Scholar, Princeton Environmental Institute and Visiting Professor, Department of Economics, Princeton University (2005-2006); Research Professor, Economic and Social Research Institute, Ireland (2006-Present); Research Fellow, Global Trade Analysis Project (GTAP), Center for Global Trade Analysis, Department of Agricultural Economics, Purdue University (2007-2010); Associate Editor, Economics E-Journal (2007-Present); Adjunct Professor, Department of Economics, Trinity College, Ireland (2010-2011); Professor of the Economics of Climate Change, Institute for Environmental Studies and Department of Spatial Economics, Vrije Universiteit Amsterdam, Netherlands (2008-Present); Professor of Economics, Department of Economics, University of Sussex, Falmer, United Kingdom (2012-Present)

Thanks to Populartechnology.net

Advertisements

  Subscribe  
newest oldest most voted
Notify of
oMan

Thanks. In lay terms, Cook cooked his books. Nice debunking by Tol.

Of course, the main failing of Cook’s study is that it is so brainless. If you want to know
what a climatologist’s current beliefs are, you ASK them, you don’t start riffling thru published works, many not even remotely current, trying to read between the lines to divine what its author’s
beliefs are, or were. Almost two decades of no warming has changed many minds, an effect that
Cook’s study implicitly claims is impossible. Asking questions about “current” consensus using
evidence that is in no way current is an oxymoron.

“[Dana] I think your sampling strategy is a load of nonsense.”
Nonsense with a purpose as well as purposely selected nonsense.

Pamela Gray

But if this had been his “Climate Science” Ph.D. dissertation, it would have passed with flying colors and put him in the running for a Nobel Prize, complete with on-line down-loadable certificate with a fill-in-th-blank space for his name and an invitation from the IPCC to have this stellar piece of sh…work…included in the next report.

ouch!

that’ll leave a mark

It’s easy to call anyone a cheat. Of course, proving it to all is quite rare, but not finishing it of by calling them out as a cheat, is even rarer still. Classy stuff Dr. Tol
Pointman

rogerknights

This paper of Cook’s is a great opportunity to counterpunch the 97% meme with an accurate survey of a representative sample. It should count only (or separately) “attribution” papers, it should poll authors about the degree of their alarmism (i.e., AGW or CAGW), and it should ask them if their degree of alarm has mitigated in recent years, as new less alarmist papers have come out.
George Mason U. should also be given a grant to conduct a new survey of members of the AGU and AMU. It’s been about six years since the last one.

Pretty impressive that Dr. Tol can shoot down the paper over Twitter. It took McIntyre et al many pages, papers, and even books to deal with the hockey stick.

Jack Savage

I think if you are going to print out Dr.Tol’s qualifications and CV it is only fair that you should do the same with Dana’s….

Steven R. Vada

People selling climate alarm
want you to believe
that 97% of people reflecting upon the following question would give answer “B”
Question: You are heating a mass. Through direct physical contact, you thermally couple a second mass, you are not heating. The temperature of mass A will subsequently go
(A) down
(B) up
===
Maybe I’m misrepresenting the 97%
Let me rephrase it:
===
You are heating a mass.
You have submerged that mass in a frigid, fluid gas bath, pulling the frigid fluid bath against the you heat, via gravity. Direct-contact thermal transfer is further enhanced by spinning the mass inside the frigid fluid gas bath.
1% of the frigid gas bath, is comprised of phase-change refrigerant. The refrigerant lifts heat faster through convection than the rest of the frigid, fluid, gas bath, dumping heat to space: makes ice – falls to the ground, to repeat the cycle.
Question: about the paragraphs above: The frigid, refrigeration-cycle augmented, gas bath,
(A) cools the mass submerged in the frigid, refrigeration-cycle augmented, gas bath
(B) warms the mass submerged in the frigid, refrigeration-cycle augmented, gas bath
In the 97%
the answer is *B*
Question 2 about the paragraph above:
The preferred analogy to the mass and it’s relationship to the frigid, refrigeration-augmented, gas bath, is:
(A) The frigid gas bath cools the mass, similarly to blowing a computer chip with a fan
(B)The frigid gas bath warms the mass, like a big, warm, blankie.
====
In magic gas country
the correct answer is *B*.
In the REST of the world
well –
– you figure it out.

Note that the above is only about the representativeness of the sample, which is poor.
I’m now looking at data quality. This is made difficult because crucial data were held back by the authors, or perhaps not collected or stored. (sic). Even with the little data available, there is convincing (to me) evidence of bias.
Oh, and there are some arithmetic errors too.

Dodgy Geezer

““[Dana] I think your sampling strategy is a load of nonsense.” – Richard Tol”
That’s a very polite way to say ‘propaganda’…

cirby

The thing about the “study” in question is how it uses key words – not substance.
For example, if someone wrote a paper about endangered frogs being run over while crossing a highway, and included the phrase “…and we expect an increase in frog mortality in future decades due to an increase in traffic to the local ‘global climate change’ laboratory,” it would still count.

klem

Wow, I wish I had a CV that looked like that.

Jim G

Richard Tol (@RichardTol) says:
June 1, 2013 at 8:31 am
Note that the above is only about the representativeness of the sample, which is poor.
“I’m now looking at data quality”.
A difficult task, no doubt, given that one must interpret the writer’s opinion where it may not be directly expressed as well as the time lag noted between the papers and present. The 97% study is patently garbage and irrelevent, in any event, given the built in bias caused by grant dollars, peer pressure and desire for survival within one’s organizational politics.

Latitude

climate abstracts from 1991–2011….
===
this is a joke, right?
You mean back when temps were going your way……before they flat lined

Skiphil

Poor, poor li’l Dana, he can’t argue the merits so now he’ s reduced to telling Richard Tol to “get some class”
https://mobile.twitter.com/RichardTol/status/340830029741187073
Sure, Dana, after you sneer and smear at everyone you are one of the last people on earth who should pontificate about who has “class”

FAH

The approach to Cook, et. al. is hindered by considering it to be an evaluation of either the “scientific record” or the views of “climate scientists.” It is much more understandable and justifiable as an evaluation of the strength of belief of those who maintain the anthropogenic, strongly C02 driven global warming belief (call them believers for now, but there are issues about the degree of belief). If the hypothesis were stated as “Do those who believe feel that the scientific evidence (or climate scientists as individuals) support the belief and at what level?” then one could say that it was a fairly good look at that hypothesis. The study as done clearly only measured how the “reviewers” used in Cook, et. al. view the scientific record. Thus the analytical path of the study simply reflected the process by which believers view the scientific record and come the their conclusion of a degree of certainty.
I fear analysis of how well or poorly the study measured something other than the belief strength of the performers is akin to the punishment of King Sisyphus. An evaluation of the strength of support in the record for even a well stated scientific hypothesis would be difficult and necessarily involve in depth analysis of the detailed content and precedence order of the work. One possible approach might be to do a trend analysis of some sort on paper subject matter to see what subjects were trending.
A more comparative study might be to design a study to survey the strength of belief of “climate scientists” themselves. But that would have to be done by interacting with the investigators and eliciting their views directly, constructing an adequate belief scale, developing a reliable elicitation process, and careful consideration of the population.
In any event, technical aspects in the Cook study that lead to significantly different measures say much more about the strength of belief of “believers” than anything else.

@FAH
That is exactly right. 3167 abstracts (out of 11944) were demonstrably misclassified. 99.4% of those were rated as endorsements.

Michael Jankowski

“In his defense, [Dana] has had limited exposure to stats at uni” – Richard Tol
That’s a pretty poor “defense.” Dana had every opportunity to involve someone who knew stats inside-and-out, right? Seems like we’ve seen this sort of issue before (repeatedly) with publications related to climate science.

thingodonta

Cook and Dana deep down believe they only need to include papers they agree with anyway, since they already know what is true. An extreme case of confirmation bias if ever there was one, just choose the papers that suit, the others are all wrong anyway.

“If your sample is not representative (which it isn’t) your conclusions are worthless.” – Richard Tol
What is Dana going to rename his blog at The Guardian when this is over?

We have issued two news releases on the Cook study via PRWeb:
“Only 65 Scientists of 12,000 Make up Alleged 97% on Climate Change”
http://www.friendsofscience.org/index.php?id=657 and,
“Friends of Science Challenge the Cook Study for Bandwagon Fear Mongering on Climate Change”
http://www.friendsofscience.org/index.php?id=655
The IPCC and climate alarmists claim that 90% to 100% of the recent warming (since 1975) was caused by greenhouse gas emissions.
Only 0.54% (65 of the 12,000 abstracts rated) suggest that humans are responsible for more than 50% of the global warming up to 2001, contrary to the alleged 97% consensus amongst scientists in the Cook et al study.
The Cook et al study data base has seven categories of rated abstracts.
1. 65 explicit endorse, >50% warming caused by man (See link in first news release)
2. 934 explicit endorse
3. 2933 implicit endorse
4. 8261 no position
5. 53 implicit reject
6. 15 explicit reject
7. 10 explicit reject, <50% warming caused by man
We can't even say that 65 abstracts in the "explicit endorse" category, or 0.54% of 12,000 abstracts, supports the IPCC consensus as many climate skeptics believe the humans (via black soot, UHI, GHG) may cause 50 to 80% of the warming, but strongly disagree that man-made greenhouse gases have caused more that 90% of the 20th century warming. There has been no warming for 16 years, a fact that strongly supports the skeptics case.

I would like to have those ‘scientists’ view this chart, and then answer the questions.

NikFromNYC

The self-destruction of climatology now has momentum, first Mann supporting a truly defective hockey stick that has no hockey stick in the input data as if it validated his complex black boxes, and now the original 97% claim is being officially propped up by fraud of such grade school level of basic fibbing that it suddenly no longer requires trusted statisticians to fully explain it all to casual readers, and the next cycle of popular books and documentaries promise to inspire an utterly massive backlash not just against rogue science but the whole Left Wing, all because they dug their own graves this fine season.

Scarface

Jack Savage says: June 1, 2013 at 7:56 am
“I think if you are going to print out Dr.Tol’s qualifications and CV it is only fair that you should do the same with Dana’s….”
There is a 97% consensus that the dog ate his resume.

Stephen Richards

“In his defense, [Dana] has had limited exposure to stats at uni”
I bet That left a mark !!

No CV online, but there is a bio:
Dana Nuccitelli is an environmental scientist at a private environmental consulting firm in the Sacramento, California area. He has a Bachelor’s Degree in astrophysics from the University of California at Berkeley, and a Master’s Degree in physics from the University of California at Davis. He has been researching climate science, economics, and solutions as a hobby since 2006, and has contributed to Skeptical Science since September, 2010.
Scopus lists 1 paper: Comment on Ocean heat content and Earth’s radiation imbalance. II. Relation to climate shifts, Physics Letters A, 2012

JJ

Richard Tol (‏@RichardTol )
@dana1981 Indeed you are. If your sample is not representative (which it isn’t) your conclusions are worthless.

In his defense, even if [Dana’s] sample is representative (which it isn’t), his conclusions are worthless.

Reich.Eschhaus

Should Cook et al have searched for “Warming” instead of “Global Warming” as well? 😉

Rud Istvan

Dr. Tol, I wonder whether you (perhaps together with others) will dig up enough information to attempt to get the paper retracted? That would be more globally helpful than the debate with Nucitelli, since could be used, for example, to counter the presidential tweet.
Appreciate your work on this.
Regards

@Rud
Dan Kammen, the editor in chief, featured in the press release.

James Smyth

:dbstealy said
Why not just use the full MSU data set?

Steven R. Vada

It’s gross incompetence to have not searched all three. With modern word processing it’s just more evidence of a stunt by a liberal to lie since he perceives all life as a lie since there are inequities intrinsic to existence.
That’s half the point : in many senses it EPITOMIZES the eco-waco/rebellion-as-yardstick ‘siyunts’
Standard physics believes in conservation of energy.
The word conservative is nearly in that, so eco-wackos are against it: and sure enough there’s Trenberth with his “More light out, than went in” energy budget – that he’s got his idiot friends in Academia endorsing.
It’s past pathetic; it’s crime really.
Teaching a generation that because there’s no policeman to imprison one for lying to get funds
it’s crime.
====
Reich.Eschhaus says:
June 1, 2013 at 12:17 pm
Should Cook et al have searched for “Warming” instead of “Global Warming” as well? 😉

Steven R. Vada

Even if it’s not a fracture of statute; which – it is and always was.

Gary Hladik

“[Dana] I think your sampling strategy is a load of nonsense.” – Richard Tol
We should give Dana some credit here: He may have worked very hard to find the search phrase that gave him exactly what he was looking for. 🙂

Rich Horton

Forgive my ignorance, but is ERL a peer reviewed journal? If so, does the word “peer” now mean “another person who does not know statistics”?

James Smyth says:
“Why not just use the full MSU data set?”
There is no established “full data set”. Your chart is arbitrary. I used the chart that I posted to show that the warming trend has been broken, beginning around 16 years ago. Since then, the planet has been cooling, even as CO2 continues to rise.
But we can go farther back if you like, to the beginning of consistent and unbroken thermometer records. The ‘full data set’ for the CET is shown here.
Note that there has been NO acceleration in [natural] global warming. The recent ≈40% rise in CO2 has NOT resulted in ANY accelertation in global warming. What is being observed now has happened before — when CO2 was much lower.
Conclusion: CO2 does not cause any measurable global warming. Therefore, the entire “carbon” scare is deconstructed. The conjecture claiming that CO2 = AGW has no scientific evidence to support it.
This is not to say that CO2 has no effect on temperature. It only points out that if there is any effect, it is so minuscule that it cannot be measured. Therefore, no additional public funds should be expended on something that is too small to measure.

But Dr. Tol is a denier! Nuccy told us so!
I guess there is still some integrity in the AGW camp. When it came time to promote the talking points or stick to science, some manage to maintain their integrity and stuck with science.

When one considers that Dr. Tol is saying that Cook, et. al. DO NOT UNDERSTAND the bell shaped curve of a “normal distribution” we then can come to a classic conclusion.
So it’s obvious that we need to say to Cook:
“Ask not for whom the TOL “bells”, the TOL BELLS FOR THEE!” (Apologies to John Donne)

Jimbo

To settle this matter I think sceptics need to produce a study then a paper based on a survey of say 200 climate scientists. It seems to me you can’t just ask one question as there will be shades of view. For example.
Q1) Has man-made co2 been responsible for most of the recent warming?
Q2) Has man-made co2 been responsible for all of the recent warming?
Q3) Has man-made co2 been responsible for less than half of the recent warming?
Q4) Has man-made co2 not had any effect at all on the recent warming?
Q5) Don’t know?
Is this suggestion sensible? Anyway, it certainly would not be 97%
See the
The Paradox of Consensus [wUWT]
“Theories that can be easily tested should have a high degree of consensus among researchers. Those involving chaotic and less testable questions – climate change or economic growth, physiology or financial markets – ought to have a greater level of scientific disagreement. Yet this is hardly the case for climate science. In the Paradox of Consensus, we illustrate that the greater the level of consensus for certain classes of hypotheses (those that are difficult to test) the less truth we should assign to them.”

Reich.Eschhaus

Q6) Has man-made co2 been responsible for more than all of the recent warming (because otherwise it would have cooled down)?

Jimbo

Would some published papers use “temperature rise” / “rise in global mean temperature” instead of “climate change” or “global warming”?

RACookPE1978

Richard Tol (@RichardTol) says:
June 1, 2013 at 8:31 am

Note that the above is only about the representativeness of the sample, which is poor.
I’m now looking at data quality. This is made difficult because crucial data were held back by the authors, or perhaps not collected or stored. (sic). Even with the little data available, there is convincing (to me) evidence of bias.
Oh, and there are some arithmetic errors too.

Thank you for the honor of your time, and the effort of your responses here.

Reich.Eschhaus says:
“Q6) Has man-made co2 been responsible for more than all of the recent warming (because otherwise it would have cooled down)?”
You need to be reminded that there is no measurable evidence showing that CO2 causes any global warming. If you believe that is wrong, present your testable, empirical evidence.

Carrick

Jimbo:

Would some published papers use “temperature rise” / “rise in global mean temperature” instead of “climate change” or “global warming”?

Probably there will be some that are missed, but that’s why it’s important to do some data quality control. For example, use authors like James Hansen who have listed all of their climate-change related pubs for validation of your search protocol. See how many you are missing and tune your search accordingly.
You shouldn’t be just getting 25% of all of Hansen’s papers or miss 33 of the 50 most cited papers for example.

tolo4zero

If you look at the abstracts Cook used, some of them actually counter skeptical sciences claim that 97% of scientists feel CO2 is causing global climate changes and the consequences could be catatstrophic.
“Atmospheric Methane – Its Contribution To Global Warming,Applied Energy,Badr| O; Probert| Sd; Ocallaghan| Pw,2,2”
Used as proof of AGW, but counters the CO2 theory, as methane is the problem.
“1991,Buying Environmental Insurance – Prospects For Trading Of Global Climate-protection Services,Climatic Change,Swisher| Jn; Masters| Gm,3,2”
“1991,Buying Greenhouse Insurance,Energy Policy,Manne| As; Richels| Rg,3,3”
Two on buying insurance ( biased maybe)
“Global Warming And The Growth Of Ice Sheets,Climate Dynamics,Ledley| Ts; Chu| Sp,2,3”
Ice sheets growing, counters the catatastrophic consequences claim.
“Varying Boreal Forest Response To Arctic Environmental Change At The Firth River| Alaska,Environmental Research Letters,Andreu-hayles| L; D’arrigo| R; Anchukaitis| Kj; Beck| Psa; Frank| D; Goetz| S,2,”
Evergreen trees at the edge of Alaska’s tundra are growing faster, suggesting that at least some forests may be adapting to a rapidly warming climate, says a new study, counters the catastrophic theory.
This is just a random look…