Professor Wei-Chyung Wang is a star scientist in the Atmospheric Sciences Research Center at the University at Albany, New York. He is a key player in the climate change debate (see his self-description here). Wang has been accused of scientific fraud. I have no inclination to "weigh in" on the topic of climate change. However the case involves issues of integrity that are at the very core of proper science. These issues are the same whether they are raised in a pharmaceutical clinical trial, in a basic science laboratory, by a climate change "denialist" or a "warmist". The case involves the hiding of data, access to data, and the proper description of "method" in science.
The case is also of interest because it provides yet another example of how *not* to create trust in a scientific misconduct investigation. It adds to the litany of cases suggesting that Universities cannot be allowed to investigate misconduct of their own star academics. The University response has so far been incoherent on its face.
Doug Keenan, the mathematician who raised the case of Wang is on the "sceptic" side of the climate change debate. He maintains that "almost by itself, the withholding of their raw data by [climate] scientists tells us that they are not scientists".
Below is my own summary of the straightforward substance of this case. I wrote to Wei-Chyung Wang, to Lynn Videka (VP at Albany, responsible for the investigation), and to John H. Reilly (a lawyer at Albany) asking for any correction or comments on the details presented below. My request was acknowledged prior to publication, but no factual correction was suggested.
Case Summary
- The allegations concern two publications. These are:
- Jones P.D., Groisman P.Y., Coughlan M., Plummer N., Wang W.-C., Karl T.R. (1990), “Assessment of urbanization effects in time series of surface air temperature over land”, Nature, 347: 169–172. (PDF here)
- Wang W.-C., Zeng Z., Karl T.R. (1990), “Urban heat islands in China”, Geophysical Research Letters, 17: 2377–2380. (PDF here)
- The publications concern temperature at a variety of measuring stations over three decades (1954-1983). Stations are denoted by name or number. A potential confounder in such research is that measuring stations may be moved to different locations at different points in time. It is clearly important that readers of publications understand the methodology, and important confounders.
- The publications make the following statements:
- (Statement A) "The stations were selected on the basis of station history: we chose those with few, if any, changes in instrumentation, location or observation times." [Jones et al.]
- (Statement B) "They were chosen based on station histories: selected stations have relatively few, if any, changes in instrumentation, location, or observation times…." [Wang et al.]
- The publications refer to a report produced jointly by the U.S. Department of Energy (DOE) and the Chinese Academy of Sciences (CAS) which details station moves, and the publications further suggest that stations with few if any moves or changes were selected on the basis of that report. However:
- Of 84 stations that were selected, Keenan found that information about only 35 are available in the DOE/CAS report
- Of those 35 stations at least half did have substantial moves (e.g 25 km). One station had five different locations during 1954–1983 as far as 41 km apart.
- If true, it appears that Statements A and B would be false, or at best grossly misleading. If false, readers would have been misled both in terms of the status of the stations and the manner in which they had been selected (or not selected).
- Keenan then communicated with the author of one of the publications (Jones) to ask about the source of location information pertaining to the other 49 stations that had not been selected using the described methodology. Jones informed Keenan that his co-author Wang had selected those stations in urban and rural China based on his "extensive knowledge of those networks".
- On 11 April 2007 Keenan E-mailed Wang, asking "How did you ensure the quality of the data?”. Wang did not answer for several weeks, but on 30 April 2007 he replied as follows:
"The discussion with Ms. Zeng last week in Beijing have re-affirmed that she used the hard copies of station histories to make sure that the selected stations for the study of urban warming in China have relatively few, if any, changes in instrumentation, location, or observation times over the study period (1954-1983)"
- Keenan points out that the “hard copies” to which Wang refers were not found by the authors of the DOE/CAS report, who had endeavored to be "comprehensive" (and that the DOE/CAS report was authored in part by Zeng, one of the co-authors on Wang). Keenan further notes that any form of comprehensive data covering these stations during the Cultural Revolution would be implausible.
- In August 2007 Keenan submitted a report to the University at Albany, alleging fraud. Wang could at that stage have made the "hard copy" details of the stations selected available to the scientific community. However, he failed to do so.
- In May 2008, the University at Albany wrote to Keenan that they had conducted an investigation and asked him to comment on it (see the rather odd letter). However they refused to show him the report of the investigation or any of the evidence to allow any comment (further odd letter).
- In August 2008 the University sent Keenan an astonishing letter of "determination" stating that they did not find that Wang had fabricated data, but that they refused to provide any investigation report or any other information at all because "the Office of Research Integrity regulations preclude discussion of any information pertaining to this case with others who were not directly involved in the investigation".
- Wang has still not made the station records available to the scientific community. If he provided such records to the University as part of a misconduct investigation, then the University has apparently concealed them.
Comments
- In the absence of any explanation to the contrary, it seems that the methodology for station selection as described in these two publications was false or at best grossly misleading.
- Wang maintains that hard copy records do exist detailing the location of stations selected by himself outwith the published methodology. However the refusal to clarify "method" is inappropriate and a form of misconduct in and of itself. It does not lend credence to Wang's assertion that fraud did not take place. It would also be necessary to see records of stations that were not selected, in order to confirm that selection was indeed random, and only "on the basis of station history".
- The University at Albany is in a difficult position.
- If the University received such records as part of the supposed misconduct investigation, then they could easily resolve the problem by making them available to the scientific community and to readers.
- If the University does not have such records then they have been complicit in misconduct and in coverup of misconduct.
- If the University at Albany does have such records, but such records are not in accordance with the stated methodology of the publications, then the University has more serious difficulties.
- "Investigations" of scientific misconduct should themselves align with the usual principles of scientific discourse (open discussion, honesty, transparency of method, public disclosure of evidence, open public analysis and public discussion and reasoning underlying any conclusion). This was not the case at the University at Albany. When you see universities reluctant to investigate things properly, it provides reasonable evidence that they really don't want to investigate things properly.
For further information on this case see here and here.
Earlier|Later|Main Page
42 comments:
Dr. Blumsohn: Thank you for this blog. I am a 1978 gruduate of Albany (math/physics/comp sci); I received an excellent education there because fraud was NOT prevalent. Some of my recent climate articles here: http://links.veronicachapman.com/Sheridan-GlobalWarmingAlarmistsExploitedIthaca.pdf and http://links.veronicachapman.com/Carbon_Tax_Questioned-h.pdf
Ethics failure in such of "success," i.e. more funding? Shameful from top-to-bottom
Paul Sheridan
UA Class of 1978
Dr. Blumsohn: Thank you for this blog. I am a 1978 gruduate of Albany (math/physics/comp sci); I received an excellent education there because fraud was NOT prevalent. Some of my recent climate articles here: http://links.veronicachapman.com/Sheridan-GlobalWarmingAlarmistsExploitedIthaca.pdf and http://links.veronicachapman.com/Carbon_Tax_Questioned-h.pdf
Ethics failure in such of "success," i.e. more funding? Shameful from top-to-bottom
Paul Sheridan
UA Class of 1978
Dr Blumsohn,
You assert that "It therefore appears that Statements A and B must be false."
Strong words.
The statements you suggest are false describe selection on the basis of "few/relatively few, if any, changes in instrumentation, location, or observation times…."
1.Do you have knowledge of the stations which were not selected because they had more inconsistencies? If not, how then do you assert falsehood?
2. Are you not aware that changes in instrumentation or observation time may be more significant than a change in location within the same micro-climate? The statements which you assert to be false generalise the 'fewness' of changes, of course - you only give examples of location change without considering the possibility that the station may have been selected for the 'fewness' of its changes in other respects.
3. Are you not aware that the examples of location change you quote from Keenan are his 'best shots' and that it is indeed true that other chosen locations show no such changes at all?
I have no particular knowledge of this case, but I suggest to that your assertions are precipitate.
There is refusal to clarify method.
The falsity of the statements depends on logic, and the way in which a reasonable reader would have interpreted them.
Unfortunately in this case there is complete inability of that reasonable person to clarify whether their understanding of meaning is correct (that the changes were small, were documented in the relevant reports and not in other reports, and that such documentation is available at all).
If I were to conduct a drug trial stating that patients were selected from the national breast cancer registry, and further selected on the basis of their female sex...
and it turned out that one or more of the participants was in fact not in that register, or was male...
then my method description would be false.
If I further refused to clarify the status of the rest of the patients, then the publication would be essentially uninterpretable as a reflection of outcome based on the described method.
Even a single station that showed a very large change, or where documentation was not available in the referenced report (or any report) would render the statements false in terms of usual understanding.
The fact that there is complete non-transparency would naturally result in suspicion that the discrepancies are far worse.
The status of the stations that were NOT selected is critical (as you suggest). That is precisely why it is necessary to know why they were selected (or not selected), and to see the relevant documentation (given that there are already documented which have defied clarification).
I for one am more interested in the mode of University behaviour - shielding an academic who declines to answer reasonable questions about published method or data (the very nature of being an academic).
Therein lies the misconduct.
Aubrey, this comment appeared here:
http://scienceblogs.com/stoat/2007/08/urbanisation_again.php
I think it is probably a fairly good analysis of what is behind this issue:
As I understand, part of the problem comes from the policy of the country C's government. It seems that they think meteorological data (except what they release real-time in the channel of WMO) as intellectual properties rather than public goods. Sometimes, international scientific agreements make some part of C's data open to public, but not always. Often, scientists who use C's data more than the open ones need to make non-disclosure agreements, formally or informally. The situation is similar to using trade secret information of private companies.
I admit that sometimes a stronger force than the non-disclosure agreement is necessary to investigate alleged frauds. It will require due legal processes. Also, in cases like this, it will also require diplomatic considerations, for it may mean that the government of one country forcefully obtains what the government of another country claims as its intelectual property.
It smells to me that the accuser takes advantage of the closed policy of C on one hand, and the principle of freedom of information of USA and UK on the other hand, to harras a scientist whose conclusions the accuser do not like.
But this allegation may be unfair to Mr. Keenan. His opinion about the issue of copyright found in his website suggests that he is a friend of knowledge as global public goods and a foe of any secrets.
I agree that scientific activity should be transparent. But it is no use to accuse the scientists who make some part of the closed information open. I hope that real promoters of liberty and transparency will join us to influence the government of C. But we should refrain from accusation, for it will make them just more arrogant. We need to persuade them with the logic that making data public and promoting global scientific cooperation is beneficial to them as well.
(I intentionally wrote "C", because the identification of the country does not matter, but just the attitude of the government.)
Kooiti Masuda
Okayeee
but that is not what he said....
If Wang reported that the stations were not in fact included in the referenced report
but were in fact included in a report that he was not at liberty to reveal
and then explained the precise criteria for selection (and non selection) -- that would allow stations with massive changes to pass selection
and provided those same parameters for stations that were not selected
then a different and somewhat more honest sort of discussion could start
Like whether the other authors on the paper saw the actual methodological data....
and if they did
then, what is the fundamental difference between me (or a misconduct committee) in terms of the government of China.
There is also the question as to whether data which is secret intellectual property (and hence not open to scrutiny) should be reported on as if it were science (which is subject to scrutiny).
...okay, I know I may not be brighest bulb on the Christmas tree but I swear I downloaded and read the pdf of Wang's article and I could't find this statement anywhere:
"They were chosen based on station histories: selected stations have relatively few, if any, changes in instrumentation, location, or observation times…."
...where exactly does that statement come from?
page 2377 Column 2
It's there
Aubrey,
I presume that you have read the papers from which the statements which Keenan attacks are extracted? The irony (given that, as you say, Keenan is a global warming 'sceptic') is that Wang's paper, for example, is full of caveats and yet concludes that UHI is actually very significant in the area (over 0.1C/decade at urban stations)! If he were engaged in fraud to support a warmist agenda, as Keenan implies, then he would surely have come up with a result more favourable to a warmist agenda!
You talk of logic, but it simply doesn't make logical sense to guess at fraudulent bias when the finding of the paper suggests a significant UHI influence.
There may well be a case to put that a phrase appearing half way through a paper could have been better worded, but that is not tantamount to fraud.
Ok, I'll accept that your concern is more to do with the university's procedure, but along the way you are contributing to questioning Wang's professional integrity. That is very serious stuff, unless you really feel confident of making a proper judgment in this matter.
So BigShittyFib turns up to tell us that its really the fault of the Mao's People's Republic of China to have a "Cultural Revolution" and a culture of extreme secrecy and paranoia against the outside world.
"During the Great Leap Forward, tens of millions of people are believed to have died, but it is not known how many tens of millions. And official records of grain harvests were often substantially exaggerated: this was not generally considered fraud, but instead making the records conform with “socialist reality”. During the Cultural Revolution, schools and universities were shut down and many intellectuals were beaten, internally exiled, or killed for being too bourgeois, and there was sometimes near-anarchy, especially in urban areas."
Not Wei-Ching Wang for fabricating data that supports BigShittyFib's favourite obsession. Because the real reason for BSF's intervention is to take attention away from the point of the fraud - the claim made in Jones 1990 that urban warming was not a serious problem in surface station records, despite all evidence to the contrary.
The only reason I can see for BSF not mentioning even the name "China" is because he doesn't want to come out and admit anything wrong with a totalitarian political system that BSF evidently admires.
And not to worry BSF - during the Cultural Revolution you'd have been in no danger.
John A.,
I see I have fans even here.
Aubrey,
This thing has been floating around the blogosphere for years and is being pumped by the climate change denialist crowd as being another indication that its all a commie hoax and al gore is fat and whatever.
Congrats for falling for it. This WILL get you an interview with Rush Limbaugh, however.
bigcitylib seems unable to make his/her own diversionary comments, and so resorts to quoting someone else's?! The notion from someone named Kooiti Masuda (quoted by bigcitylib) that misconduct lies ONLY with UofA is absurd.
The Simon Evans comment is also diversionary (at best); his pontifications about "knowledge of the stations" and "changes in instrumentation or observation time" and "examples of location change," etc. have NO RELEVANCE to the narrow ethical/misconduct subjects at issue. Nor do regurgitations about Keenan as a "skeptic" (ostensibly therefore I agree that-that qualification by Dr. Blumsohm, et al. should be omitted).
The whole, and rather simple blatant point, is that various climate esoterica (for some) are in-question (no matter what one claims to be or not be "aware" of). Indeed you, me, and the volcanoes on Venus might be universally recognized experts on this esotercia, but NONE of that is relevant.
Alternatively, a blatant admission does emerge from the convoluted comments of bigcitylib/Masuda; that is, if China or wherever is not "open" and has no "transparency" than how can that EVEN REMOTELY QUALIFY as a basis for the Wang publication(s)?? If transparency cannot be had with some portion of the China data then it must by-definition be abandoned. Are bigcitylib/Masuda demanding intrinsically that the rest of us believe in the new "trust me" method of climate science? Leave my name off that list.
Science is about transparency, precision/reliability and integrity/ethics. Whether you agree/disgree with the resulting this-or-that climate science opinon/esoterica (Wang's or Keenan's) is NOT relevant to this case.
Indeed, if you are going to indulge in intellectual bigotry then, in that context, some may pronounce that the behavior of UofA proves that it is merely a fundraiser, not a trier/seeker of truth. As you can see, it gets us nowhere.
Stick to the issue . . .
Yes, BSF we know that your prime object is to divert attention away from science and the conduct of science by making baseless claims attacking Aubrey's beliefs. Which is all you ever do.
The only reason why you would attack Aubrey's ethics is because you don't have any.
Whether other people think AGW is or is not a problem, scientific ethics are not optional when it comes to establishing the validity of scientific claims. Aubrey knows it and most credible scientists do.
I have known about Wei-Yung Chang for some considerable time, because it was Steve McIntyre who first drew attention to the extraordinary claims made in such a key paper. But knowing that something is amiss is very different being able to prove that something is amiss.
At the very least, Professor Chang is may have committed scientific misconduct. At the very least SUNY is attempting to cover up that fact, breaking its own rules and federal law in the process.
Just expect BSF to throw more mud, because that's what BSF does. Just expect him to mention "China".
I posted this to the comments on WUWT:
I think you are all missing something.
Read the odd letter as a lawyer rather than a layman and it all makes sense.
The letter says the “the investigation committee finds no evidence of the alleged fabrication of results and nothing that rises to the level of research misconduct having been committed by Dr. Wang.”
The operative phrases in this two part sentence are (1) fabrication of RESULTS and (2) having been COMMITTED BY DR. WANG.
The way I read the letter - and I have written many similar ones in my career - is that Albany found something. And it was large and was research misconduct. It involved the fabrication of DATA not results and it involved research misconduct by someone other than Dr. Wang. I have my suspicions of whom they are referring to but they are just suspicions.
This letter is true, accurate and very deceptive. It covers Albany’s posterior while not actually lying.
I would venture to guess that Albany has a definition of “research misconduct” that does not include failing to supervise a grad student or failing to thoroughly check the work of your co-author.
This would also explain why Dr. Keegan was not given the report and why he was not interviewed.
My guess is that the preliminary inquiry quickly established that real malfeasance had happened on this file but that it was not directly attributable to Wang.
The Investigation Committee investigated Wang and someone else. Found the someone else culpable. As a result it never needed to go back to Keegan and would legitimately have to keep the report from Keegan because it deals with a misconduct that is different than the one he reported.
Always read these letters looking for weasel words and deceitful language and you will likely get closer to the truth.
Some access to information requests on the other authors and other investigations on this paper may be in order.
Just my wild speculation.
BigCityLib,
Imagine an article on a new drug therapy that said, "...cases were randomly selected from a private proprietary database maintained by the drug manufacturer." No journal would allow that to be published, and no reviewer would accept the article's conclusion that the drug was was the great success that was claimed in the article.
So instead, imagine the article said "...cases were randomly drawn from the NIH open access patient registry," when in fact they were drawn from the secret proprietary database. That is research misconduct, and would almost certainly result in the journal publishing a retraction, and at least the possibility of public agencies withholding future funding.
It is vitally important that original data be available for inspection, and in climate science you do not even have the patient confidentiality barrier to overcome.
So with Dr. Wang the base case scenario (according to you) is that he used data from a private database that he can not release, and that no one one else can inspect, and that he misrepresented his source and methods. Either that is scientific misconduct or it is not science. (And am I the only one reminded of microfilmed probate records that were notated on yellow pads that were subsequently destroyed in a flood, and records that were destroyed in the San Francisco earthquake and fire of 1906 subsequently turning up in a small public library that has no record of the researcher's visits or the corresponding documents even existing?)
The original author might want to contact the journals that published this material.
Quote: "Imagine an article on a new drug therapy that said, "...cases were randomly selected from a private proprietary database maintained by the drug manufacturer." No journal would allow that to be published, and no reviewer would accept the article's conclusion that the drug was was the great success that was claimed in the article."
Herein lies the irony.
Most drug trials *are* in fact based entirely on "private proprietary databases maintained by the drug manufacturer".
We have bent ethics in that domain to accommodate vast amounts of money and the vested interests of some big players. The consequences have been dire, with numerous instances of fraud and deception. The outcome has been similarly dire for patients, the long-term health of this industry, small science based companies, and the credibility of scientific medicine and of "regulators".
So whether it involves pharmaceutical "proprietary" secret-science, or secret Chinese data, the principles are, or should be, the same. Science is open to scrutiny or it is not science at all.
Hiding stuff simply makes the whole thing outside-science, and the only beneficiaries are those who don't want a scientific answer, but instead a
political one
or a magical one.
Re comments by Simon Evans.
I effectively made two allegations. One is that there is no data for 49 stations; the other is that the data for the remaining 35 stations does not conform to the quoted statements by Jones et al. and Wang et al. The first allegation is obviously the more serious one, and yet your comments ignore it. The second allegation is indeed subject to some interpretation, but the descriptions on my web site (including maps) make it clear that the data from many of those 35 stations do not conform to the statements.
Regarding the first allegation, the evidence for it, detailed in my report to the university, appears conclusive. Moreover, the university conducted a preliminary inquiry, for which a (redacted) report is available. The report states that "there were a sizable number of stations for which there was no data".
As far as motivations for the fraud, I do not know what was in Wang's mind. Perhaps, like almost all researchers, Wang wanted prestigious publication.
Douglas J. Keenan said ... (May 04, 2009) :
Re comments by Simon Evans.
I effectively made two allegations. One is that there is no data for 49 stations; the other is that the data for the remaining 35 stations does not conform to the quoted statements by Jones et al. and Wang et al. The first allegation is obviously the more serious one, and yet your comments ignore it.No data? Are you now insinuating that the data was entirely fabricated? I will guess that you mean the data has not been made available. You may question whether that is proper or not, but it is not prima facie evidence of fraud.
The second allegation is indeed subject to some interpretation, but the descriptions on my web site (including maps) make it clear that the data from many of those 35 stations do not conform to the statements.Yes indeed, it is 'subject to some interpretation' but no, your descriptions do not make your allegation 'clear'. You are applying your presumption of what 'few' should mean. I think you would have a good point if you were saying that the papers failed to clarify their terms, but that would not be fraud (or fabrication).
Regarding the first allegation, the evidence for it, detailed in my report to the university, appears conclusive.What, that there was never any data? Where is your 'conclusive' evidence of that? You can keep repeating that you consider your evidence conclusive, but repetition doesn't make it so.
Moreover, the university conducted a preliminary inquiry, for which a (redacted) report is available. The report states that "there were a sizable number of stations for which there was no data".Well, Dr Keenan, that is a remarkable piece of selective quotation on your part. What the report says is:
"The Complainant highlighted points that the Committee had taken notice of, namely that there were a sizable number of stations for which there was no data that could be used in defending the statement that these stations had remained relatively "constant".
The Committee had taken notice of your point, it did not make a determination of whether it was valid or not, as you imply it did.
In thus manipulating a partial quotation you are engaing in misrepresentation. You take this further in your 'timeline' notes -
The Committee unanimously concluded that “there was no data” (thus implicitly concluding Wang must have fabricated data) and that a full investigation should be undertaken. (my bold).
The Committee concluded no such thing. Your statement is shocking distortion.
"it seems that the methodology for station selection as described in these two publications was false or at best grossly misleading."Indeed; it seems likely. I note that you have not used Keenan's word (fraud). I presume that is because the evidence in the public domain goes nowhere near substantiating that charge against a named individual.
"If the University does not have such records then they have been complicit in misconduct and in coverup of misconduct."I do not see that this analysis can hold. Why should the UoA have these records, 19-20 years after the publication ? Current UK research council guidelines only suggest holding original data for seven years.
If the UoA does not have the original data that it needs to investigate the fraud charge, how can it make an adverse finding ? Surely that is reasonable doubt ?
It may be that Douglas Keenan can make headway on procedural grounds; certainly the description of procedure does sound strange. I will be fascinated to see.
per
Dr Blumsohn
you refer to a climate change “denialist” and also say that Doug Keenan, the mathematician who raised the case of Wang is on the “sceptic” side of the climate change debate.
Because the alarmists, for obvious reasons, have changed “Anthropogenic Global Warming (AGW)” to “Climate Change,” they now incorrectly call those of us who oppose AGW and “Climate Change Sceptics” or “Climate Change Deniers.”
As Climate has always changed, I defy anyone to point out an Anthropogenic Global Warming Sceptic who denies that Climate Changes.
Having got that off my chest, it is interesting that although the papers have several co-authors and on one he wasn’t even the lead author, only Professor Wei-Chyung Wang is the only one accused of fraud.
The paper authors were
1 Jones P.D., Groisman P.Y., Coughlan M., Plummer N., Wang W.-C., Karl T.R.
2 Wang W.-C., Zeng Z., Karl T.R.
Surely if one is guilty of fraud, all are.
And they join a long list of AGW alarmists guilty of corrupting data.
This anonymous "per" of May 04, 2009 apparently suffers from . . . something. Get a load of this convoluted nonsense: "Why should the UoA have these records, 19-20 years after the publication?"
In other words, just formulate policy and spend TRILLIONS chasing science that cannot be reproduced, subvert the scientifc method to that of being faith-based? Is that the per ploy? Forget it. And what makes this anonymous "per" think his/her CLAIM that "Current UK research council guidelines only suggest holding original data for seven years," has ANY relevance?
But this mindless "per" rant takes the proverbial cake: "If the UoA does not have the original data that it needs to investigate the fraud charge, how can it make an adverse finding? Surely that is reasonable doubt?"
Where do these people come from??
The comment by DublD has been removed. Any reasonable comments accepted (with further leeway if the author is identifiable), but this anonymous comment went beyond that into the realms of ad-hominem and possible racism.
There are legitimate questions here, like:
"How long academics should keep data or methodological information" (in my view it is reasonable to think that this would depend on the type of data - I imagine that information about such stations would be expected to be kept pretty much indefinitely - if it existed - given the likelihood it would be required for subsequent reports).
I guess a proper report from UaA would generate clarity
Ed
This is only the tip of the climate iceberg (sorry couldn't resist!).
The fact that the raw data is hidden from the scientific community as a whole is a breach of every scientific principle and every scientist of any integrity should be screaming "fraud" until such time as the complete raw data set is routinely released for scrutiny.
Sure
this is correct.
However it applies equally to many aspects of commercial "science" (particularly pharmaceutical) where there is no such screaming - just quiet collusion of journals and regulators.
Drug companies can be forced by legal means to release the data (I know because as an infectious disease MD I have been there!).
But there is no patentable product based around temperature data and that data is also procured with public funding. There is no legal, financial or ethical reason for withholding it!
Actually I disagree slightly. there are almost no instances of drug companies releasing actual raw data into the public domain (usually if we are lucky they release their own proprietary version of what they say the data shows).
The fact that there is a patented product is irrelevant. It depends what data we are talking about. If we are talking (for example) about their magic manufacturing method that could certainly be a secret.
However if we are talking about how the drug works, that cannot possibly be held to be a secret. The product they are selling IS a package deal of the molecule and open honest science. If there is no open science there is no product (and the molecule should not be sold under the banner of science). It could be sold under the banner of quackery (as are homeopathic products) I guess.
""How long academics should keep data or methodological information" (in my view it is reasonable to think that this would depend on the type of data - I imagine that information about such stations would be expected to be kept pretty much indefinitely - if it existed - given the likelihood it would be required for subsequent reports).
i think there is a difficulty, insofar as you cannot tell in advance when a study is going to be something that other people rely on. You seem to be making the case that Universities should maintain all raw data without time limit. I have to point out that there are practical difficulties in that scenario.
"The fact that the raw data is hidden from the scientific community as a whole is a breach of every scientific principle and every scientist of any integrity should be screaming "fraud" until such time as the complete raw data set is routinely released for scrutiny."
All due respect, but I do not think that this is a reasonable perspective. While some science (e.g. large scale physics projects) make all their raw data available, once you get to the scale of lab-based work, I think it would be very unusual to request the raw data (notebooks ? patient records ? computer files ?), never mind provide access to. That is the stuff of formal misconduct investigation, and it is the responsibility of the institution by and large.
I do not think that you are accurately portraying a reasonable perspective for the vast majority of published research.
Finally, i should add that I feel honoured, 'touched by the hand of god'...
From Veronica Sheridan's web site
"Consequently it is possible to argue that real Physics has been suppressed, declared ‘top secret’ in order to maintain dependence on oil, to the advantage of those promulgating the suppression."
I feel special to take abuse from her.
per
Anyone capable of thinking beyond their nose can see that the bigger issue here is that China has a significant land mass. It has a significant population, several large cities and gets most of it's power from burning coal and wood.
Now if what I'm reading is correct, there is a 30 year period of temperature data that is at BEST highly suspect, at worst fabricated or non-existent. So the implications of this on every "Global Mean Temperature" record in existence are profound. Does this negate what many have observed as Global Warming? Does it compound it? Or more importantly "Can we even know at all?" It would seem to me that that answer is no.
So regardless if Dr Wang is guilty of misconduct or not (though it's pretty obvious he's hiding something), the bigger more certain conclusion that can be drawn from this is that "the science of Global Warming is NOT settled", and probably cannot be in the manner which many thought it was. And while the media will try like hell to suppress this, at least now, those of you who have read this article, will be privy to the manner in which they profess as fact that which is the furthest thing from it.
So while the extent of this evidence may not totally make Dr Wang a fraud, it sure makes Jim Hansen, Al Gore, Joe Biden and Barry O big, fat liars. There is now officially an acknowledged flaw in the data supporting any assertion that Global Warming is real and happening. Think they'll stop pushing it?
To the person who sent me an abusive E-mail:
I am happy to post just about any comment on this blog so long as it is made in the name of a real, identifiable person with a real E-mail address who bears responsibility for what they say (just as we expect honest scientists to bear responsibility for what they say).
Anonymous comments I will moderate at my discretion and generally with a very light touch if at all relevant. My discretion reviles against posts that imply that all Chinese persons are fraudsters.
There is some irony in the use of anonymous abuse while arguing for accountability.
Simon Evans, replying to Douglas J. Keenan, says "Are you now insinuating that the data was entirely fabricated?" He could only ask such a question if he hadn't read Keenan's report to the university. So he is being strongly critical without having even looked at the four-page document that he is being critical of.
http://www.informath.org/WCWF07a.pdf
His "interpretation" of the paragraph from the Committee is at manifest distance from reality.
What would motivate him to make such postings?
"Actually I disagree slightly. there are almost no instances of drug companies releasing actual raw data into the public domain (usually if we are lucky they release their own proprietary version of what they say the data shows)."I was involved in two court proceedings where a certain pharma company was forced to release its raw data on adverse effects. Patients were identified by a number only and AFAIK these court proceedings were not sealed.
"The fact that there is a patented product is irrelevant. It depends what data we are talking about. If we are talking (for example) about their magic manufacturing method that could certainly be a secret."But the patent MUST disclose the drug and it's uses and is the protection from copying the drug. What is there to "protect" in the raw temperature data because there is no possible financial benefit or patentable product.
"All due respect, but I do not think that this is a reasonable perspective. While some science (e.g. large scale physics projects) make all their raw data available, once you get to the scale of lab-based work, I think it would be very unusual to request the raw data...."This is not private or personal data, rather it is the raw temperature readings form all the monitoring stations identified by location. This is the raw data which is "fudged/corrected" to give the impression that the earth has warmed. And yes when these scientists are using this data to press for change to the very fabric of every society on earth you bet they should release the data upon which they are basing these demands so it can be scrutinised by every man and his dog!
Anonymous said ... (May 06, 2009) :
Simon Evans, replying to Douglas J. Keenan, says "Are you now insinuating that the data was entirely fabricated?" He could only ask such a question if he hadn't read Keenan's report to the university. So he is being strongly critical without having even looked at the four-page document that he is being critical of.Keenan's report asserts that statements were fabricated. I have asked him if he is now asserting that the data was entirely fabricated. Are you able to understand the difference?
His "interpretation" of the paragraph from the Committee is at manifest distance from reality.You think so. I don't.
What would motivate him to make such postings?Well, 'Anonymous', firstly I'll point out that I am posting under my own name and the website owner has my email address. Keenan has accused Wang of fraud, unreasonably in my view, and continues thus to acccuse him in public forums following Wang's exoneration by the inquiry. I think that is improper behaviour. I have also made clear that I consider Keenan has misrepresented the University's statements by means of partial quotation and the attribution of a judgment which was not given. I think that is also improper. You may conclude, therefore, that my motive in making these statements is the same motive as that which I must presume to have been Keenan's motive.
Geoff Brown, I agree that, as a general principle, all authors should take joint responsibility for the content of a publication. In fact, I think some journals now require all authors to sign a document saying that they agree with all the results.
The paper of Jones et al., though, is an exception, I think. The paper analyzes data from three regions: China, Russia, and Australia. Wang supplied the data for China; other co-authors supplied the data for other regions. Jones then did the analyses, substantially relying on methods developed in earlier work by Karl. (Jones actually drafted the paper, then sent it out to each of the other co-authors for comments, then revised, etc.)
Each data-supplying author was trusted to supply good data, and other authors did not independently check that. Indeed, such checking would have been difficult, because Wang's records were partially in Chinese at the time, and some of the other records were presumably partially in Russian. So, for the paper of Jones et al., I think that Wang gets all the responsibility for the Chinese records.
I am not sure about co-author Zeng in the paper of Wang et al. Karl would have just trusted Wang.
By 2001, though, Jones knew that there were problems with Wang's work. Yet Jones continued citing that work in his own papers. This is discussed in section 2.4 of my paper in Energy & Environment.
I see that Keenan is not a man to respond to direct questions, so I will not be following this thread any further - no responses here to any further comments.
The email collection recently released/hacked from CRU seems legitimate so far, and it contains something relevant to your post on the Wang case.
It appears in one of the emails that Tom Wigley (Director of CRU) was also quite concerned. He "would not be surprised if [Wang] screwed up here."
He shares your concern about the handling by Albany SUNY.
The rest of the document, especially Wigley's comments at the phrase "my views," is quite interesting.
http://www.anelegantchaos.org/cru/emails.php?eid=972&filename=1241415427.txt
Missing from the collection of emails is the reply to Wigley's concerns about the person he describes as a "sloppy scientist."
===|==============/ Level Head
Leaving aside the question of fraud, this case seems to expose another instance of research where the original data is not available.
The scientific pursuit of truth is supposed to be an open and verifiable process.
I'm not sure how a scientist expects to defend their work if they don't make the data, processes, procedures, and software that they used to create that research available for inspection by any interested party.
Chris Shaker
This is all very interesting and may take some time to figure out.
There can very well be scientific misconduct without actually the conclusion being wrong.
There is a big assumption being made anyway to do this study at all - that the *average* weather at a rural site geographically near (within a hundred miles or so) to an urban site will be similar and that whatever differences remain are due to the urban heat island effect.
Even if the urban or the rural site was moved that would still be the same - that is the misconduct might not effect the validity of the results. You'd just have more pairs of sites than mentioned. (each different urban/rural comparison being a different pair)
It just could mean that the reason for picking them is a lie, or a vast oversimplification that hid some facts. But the reason for the lie might not be an intent to commit fraud, that is to give invalid results - either on the part of Wang or some person or persons in China.
But this is where there is grounds for suspicion - because in effect we have been told this is a random selection - the data selected being the best they can get. It all depends on whether *that* implied assertion is the truth.
The fact that what he said is not the truth doesn't mean that the sites were in fact cherry picked - they might have been selected for the non-interruption of weather records, which is what he says, except that it is omitted that there was a lot of interruption of the gathering of weather data in China during those years and that he really has many more "pairs" than he admits.
We don't actually know what went on of course.
It is clear of course that something is being hidden. For instance, why does this start with 1954?
I thought first that maybe that would be a time when the Communists had full control and there were no more arrests being made or schools or institutions being closed.
But then I had a better idea.
Someone raised the question of how could the data be maintained through the Cultural Revolution. and then it hit me.
Maybe they were not lying about the records being maintained from 1966 through 1978. It might be that many of these sites - especially the more rural ones - were maintained by the Chinese military, and that would not have been interfered with. And it would have taken them to 1954 to really get going.
Of course it is quite implausible to believe that there were no rural stations till 1954. It's just that many of the old ones were closed. And of course they don't want to say that.
These stations would have been established for weather forecasts for military flights and later for atomic tests. They might be near or at secret airfields or other military installations. This is probably part of the reason for the secrecy.
If the stuidy is nt a total fraud in fact these stations actually practically have to be military weather records. Only if these staions were maintained by the Chinese military would there be few before 1954, major interruptions for others surrounding the Communist takeover of power but no interruptions because of the Cultural Revolution.
Some might also be at prisons or slave labor camps since they might be the only convenient somewhat built up area and it is easy to add another duty to a place. That would constitute even more reason for secrecy as to the locations of these weather stations.
Besides which, of course, everything is a secret in China unless there is what strikes the government a good reason for disclosure. But more likely they have some specific reason for the secrecy.
Douglas J. Keenan wrote in The Limehouse Cut, London E14 6N, UK;
03 August 2007.
"It is clear that when a station moves, the temperature data from before the move is not, in
general, directly comparable to the data from after the move."
This can very well be true - and can raise questions about other matters - but in the 1990 study, Wang was not comparing data at one station with earlier and later data at the same station, but was comparing data from an urban station with data at a rural station collected at the same time!!
All that does is create more, equally valid, pairs of weather stations.
For the purposes of the 1990 study, actually, the MORE pairs of data you've got - the more the data is split up into separate pairs - the better the data is.
It doesn't matter too much WHERE in a urban area a station is, and it doesn't matter much where a rural station is. Just so long as it's still a valid pairing, and so long as it wasn't variations in temperature readings that caused the movement of the station.
Which actually might be a reason for moving the station, come to think of it. But this would most often happen when it was something awry within a few hundred yards or less.
I think the effect would be to minimize the heat island effect of official data (since urban data is more readily artiicially hot tha cold)
The question is, though are the data true or being fabricated and
is anybody biasing the results.
You need to know there was no bias in the selection of stations to compare. It all depends on what might have been the intentions of the government or local people in China in participating in this study.
By the way, I found the Anonymous posting of May 04, 2009 very insightful. In fact this kind of thing *sounds* like Albany or sometimes Washington..
Anonymous said:
Begin quote:
------------------
"The letter says the “the investigation committee finds no evidence of the alleged
fabrication of results and nothing that rises to the level of research misconduct having
been committed by Dr. Wang.”
The operative phrases in this two part sentence are (1) fabrication of RESULTS and (2) having been COMMITTED BY DR. WANG.
The way I read the letter - and I have written many similar ones in my career - is that Albany found something. And it was large and was research misconduct.
It involved the fabrication of DATA not results and it involved research misconduct by someone other than Dr. Wang."
------------------------
End of quote.
There is another interpretation maybe to put on this. Maybe teh most important words are NO EVIDENCE and RISES TO THE LEVEL and RESEARCH.
Wang wrote first - for his TIME FRAME - that it was chosen "mainly" because most stations (that persisted into the 1980s?) were established by 1954and "continuous records exist".
BTW, Why the "mainly?" This requires some thought. What other reasons could there be? What is he leaving out?
Then he wrote:
"They [the weather stations used] were chosen based on station histories: selected stations have relatively few, if any, changes in instrumentation, location, or observation times…."
Now while that might be untrue, it
would not be fabrication of data, and it wouldn't "rise to the level" of "research" misconduct.
It could be a lie, but the investigation committee concluded it was not RESEARCH misconduct, but only REPORTING misconduct..if it didn't change the results, and they concluded it didn't.
But maybe it is not even a lie:
Note the word "relatively"
Relative to other stations they might have picked. And some had no changes.
That statement could also just maybe be misleading, in that people don't realize how much inconstantcy there was with weather data collection in China, something that the government of China is not interested in enlightening people about.
The data as to location etc, especially about the rural stations, is a state secret in China, but it still could be true, although you'd have to rely on somebody in China to tell you which station records to use, and there's every indication that they did. The work is not checkable.
And you have to assess whether or not the government of China might be interested in fabricating data.
This is art, not science, but it may be a valid way of learning things.
Hopefully you guess correctly when the government of China is engaged in statistical lying and when it is not. They may not have been lying here.
By the way, when China has a secret - the fact that something is secret is ALSO a secret. One of the biggest secrets for instance, is just what is being censored on the Internet. (In this China differs from Saudi Arabia)
So Wang can't say anything without compromising his working relationship with people in China.
Now not revealing that Wang et al only saw the data that China passed on (according to criteria they gave China) may be misconduct because maybe China didn't follow instructions.
Sammy Finkelman says,
"in the 1990 study, Wang was not comparing data at one station with earlier and later data at the same station, but was comparing data from an urban station with data at a rural station collected at the same time!!"
Please read the paper more carefully, Sammy. The entire purpose of the study is to examine trends among stations.
A trend is a time series. Mess with the underlying metadata of your data source, and you've got a problem.
In line with the general theme of this site: suppose a medical study used data from subjects who switched from ingesting the prescribed amount, to half-doses. And others slowly increased their intake until it had doubled. But none of this was recorded.
Without the underlying metadata it would be a bit difficult to conclude much of anything, eh?
Unfortunately, I've seen way too many climate studies in my time, that select the study data on the basis of the already-recorded data values (ie. how well the data values correlate with one another, or how well they match a reference temperature slope), rather than the underlying metadata. Ie, criteria are too-often specified as a function of data content rather than context. That's not good science, not something that should be accepted uncritically.
Simply put: the selection criteria should be described well enough that a future study could be performed using the same criteria, without peeking at the results.
Sadly, that is incredibly rare in this corner of the scientific endeavor.
The February 2010 submission of Benny Peiser to the Parliamentary Inquiry (Science and Technology Committee) is here (#38). A better-formatted version of Peiser's memorandum is available here.
A 3-page PDF document which appears to be the unredacted version of the SUNY-Albany "Report of the Inquiry Committee" was among the files leaked/hacked/stolen from UEA and released in November 2009. Douglas Keenan has archived this PDF at his website. It is linked under "2009-11-17" at this page.
Post a Comment