Science the victim of dishonest attacks

That’s the title of my Fin column for Thursday 11 March 2010, which naturally picked out The Australian newspaper as a prime vehicle for these attacks. The Oz replied next day, with characteristic mendacity, pointing out that, on the same day they

ran an opinion piece by climatologist James Hansen, the NASA Goddard Institute for Space Studies chief who also happens to be known rather snappily as the “father of global warming”.

Only problem was, they weren’t running Hansen to defend science against their attacks, but because his policy views (he opposes an ETS and supports nuclear power) could be used in their continuing wedge campaign. The piece (can’t find it to link ran under the headline “”Only carbon tax and nuclear power can save us”

Anyway, here’s my piece

Science the victim of dishonest attacks

It is a commonplace to observe that Australia’s scientific institutions and organizations, have played a central role in promoting Australia’s prosperity and in maintaining our country’s place as a leading contributor to the growth of knowledge.

In city and country alike, we rely on the predictions and analysis of the Bureau of Meteorology, predictions that have grown steadily more accurate over time. The prosperity of our rural sector has been built to a substantial extent, on the work of the CSIRO and other organizations devoted to agricultural science and natural resource management.

Universities have also played a crucial role. My own University of Queensland includes among its alumni such great scientists as Peter Doherty, whose work on immunology won him the Nobel Prize for Medicine in 1996.

In recent years, science and scientific institutions have come under increasingly vociferous attack, with accusations of fraud, incompetence and even aspirations to world domination becoming commonplace. These attacks have mostly focused on environmental and public health issues, but they are gradually coalescing into an attack on science itself

A few examples

* In November 2003, Quadrant magazine published an article by Ted Lapkin blaming environmentals scientists for a supposed ban on DDT that had, he claimed cost millions of lives. DDT was never banned in anti-malarial use, and the claim Lapkin repeated had been cooked up by a tobacco lobbyist, who sought to put pressure on the World Health Organization, then campaigning against smoking in the Third World.

* On March 5 2006, Miranda Devine wrote that ‘Environmentalism is the powerful new secular religion and politically correct scientists are its high priests … It used to be men in purple robes who controlled us. Soon it will be men in white lab coats. The geeks shall inherit the earth.’

* On March 26th 2009, Jennifer Marohasy, then a Senior Fellow at the Institute of Public Affairs, accused the Bureau of Meteorology of tampering with weather data to fake evidence of global warming

* Andrew Bolt of the Sun-Herald has repeatedly asserted that climate scientists are conscious frauds, motivated by a desire for government grant money, most recently a few days ago in a blog post entitled ‘That buys a lot of Baas’.

* The Australian newspaper has campaigned against science and scientists so consistently that picking a single example would be misleading. Blogger Tim Lambert, who maintains a running series on The Australian’s War on Science is now up to instalment 46

All of this has reached a crescendo in the wake of the so-called Climategate affair in which a group of ‘sceptics’ harassed climate scientists at the University of East Anglia with a campaign of deliberately vexatious form-letter Freedom of Information demands, hacked the University’s email system to obtain the email files generated in response and then published distorted versions of those supposedly proving that global temperature records had been fudged in a ‘trick’ to ‘hide the decline’. Subsequent inquiries showed that the selectively quoted phrases referred to perfectly legitimate methods of data analysis, but the enemies of science had a win in the media.

Scientists have been constrained in fighting back by the fact that they are ethically constrained to be honest, whereas their opponents lie without any compunction. A striking example was the response of Phil Jones, the main target of the Climategate hack, when presented with deliberately loaded question about the statistical significance of global warming trends over short periods.

Jones answered honestly, and proceeded to explain the problem with this kind of analysis. The Daily Mail promptly ran a headline stating ‘Climategate U-turn as scientist at centre of row admits: There has been no global warming since 1995’

As the Economist observed, this was a flat-out lie, noting that ‘anyone who has even a passing high-school familiarity with statistics should understand the difference’ That did not stop dozens of anti-science commentators from passing it on.

Now, however, science is pushing back, at least in Australia. Along with other scientific institutions, Universities Australia is organizing a national policy forum on climate change to be held in Parliament House next week which will not only restate the findings of science on this issue but respond to the stream of attacks on science.

Australia can, if need be, do very well without Quadrant, the Institute of Public Affairs and The Australian. We cannot do without science and scientists. The time has come to make a choice.

230 thoughts on “Science the victim of dishonest attacks

  1. @Neil Fisher

    So what’s the issue with waiting 10 years if a century of warming is overwhelmed by such noise?

    By the way, just because some noise exceeds the signal, doesn’t mean there’s no signal. The noise in this case (weather) is high frequency, the signal (global warming) is low frequency. Not difficult to separate the two.

  2. @Chris O’Neill
    Chris – I think the real agenda of Mr Fisher is summed up in his own words

    ” Would you like to learn in 10 years time that you have paid 100% more tax than you needed to because someone forgot to carry the one? ”

    Its not about the science – he is one of the “lower taxes” whingers. Where he gets the “100% more tax” from is anyone’s guess but there is nothing scientific about that number so why should the rest of his so called numbers be accurate (and indeed you are correctly pointing that out to Mr Fisher)..

  3. @Chris O’Neill

    Why is it difficult to accept post-hoc adjustments made to the base data, assuming they are as large as you claim?

    I do not have a difficulty with pos-hoc adjustments to the data – the difficulty I have is that such adjustments do not appear to be reflected in the uncertainty intervals when they clearly should.
    And they are that large – a fact you can check for yourself at the NCDC, NOAA and GISS web sites. I will not guarentee that they are exactly what I have said, but they are certainly close to it – as I noted at the time, those figures were from memory. As far as I am aware, there is no dispute about this fact in climate science circles.

    If you don’t like the data from the early 20th C then just use the data from 1964.

    Once again you misrepresent my point – which is that even if such adjustments are required and fully justified (IMO, they are not, but that is another arguement) I can see no reason why the adjusted temperature for 1930 (say) should be different in 1990 than it is in 2010. If the method of adjustment hasn’t changed (it hasn’t changed significantly according to those who make the adjustments) then data from the current time period should in no way affect the measurements made in 1930, nor should it make a difference to any adjustments made to that data to account for known (or suspected) problems. The fact though, is that this does happen. I can imagine most peoples reaction if the tax office used a similar method and send you a letter saying something like “Based on your 2010 tax return, we have determined that we need to apply a correction to your 1980 taxable income and after such correction is made, the following amount is now due, which includes interest and penalties”. You’d hit the roof and rightly so. Yet you appear to accept that this is correct and proper for GMST.

  4. @Neil Fisher

    That isn’t how science works at all. That is how religion works. Science doesn’t work by an audit process. Science works by people going out and doing their own research. Not through critical examinations of ‘sacred’ texts. And who is McIntyre to set himself up as an auditor? I am not going to audit McIntyre. Who has sufficient time to audit every crank? Climate Audit is nothing but an exercise in ad-hom. It has no other objective than ad-hom. That is what the auditing is all about. If the climate scientist are wrong. Go out do your own research. Where some scientist makes a claim, like ‘cold fusion’, and it is nonsense, they are soon found out. And they are found out not by any audit. They are found out because no one else is able to get what they got. Their results cannot be replicated. If AGW deniers think that everything can be explained without anthropogenic greenhouse gases let them do their research and so it. That is science. What climate audit is about is non-science by a non-scientist. The fashionable science of science by ‘press release’ as opposed to peer review. The sort of science that would take someone as being a great scientist simply because they are a presenter on TV or write a few popular books.

  5. No one is stopping McIntyre from doing real research in climate science. That he choose to not even try says a lot.

  6. @Alice
    Alice, the details do matter – from the point of view of science, we should be correcting any errors found regardless of their magnitude or direction; from the point of view of public policy involving trillions of dollars of public money, even a small error can easily amount to millions of dollars, and people are regularly brought to account over such amounts of public money being spend by govt departments. For example, many people bemoan that federal govt budgets are out by $1 billion dollars, even though this represents less than 1% of the budget – getting within 1% is pretty bloody good for projections of economic activity up to a year in the future. If you doubt this, perhaps you should project your household budget 1 year into the future and keep real income and expenditure within 1%. I think you’ll find it’s difficult. Nevertheless, this is what happens – people do moan about such piddling amounts. Given that this is true for health, transport and many other govt functions, and given that even those who should know better complain about such things, is it too much for me to jump in and demand the same from cliamte policy? I don’t see that it is, because the same arguements apply here as they do to those other areas.

  7. @Alice

    Where he gets the “100% more tax” from is anyone’s guess

    Quite right Alice, and passing over the fact that an ETS is not a tax but a fee for service — the right to emit CO2 — it is also compensated in cash or kind for the vast majority — anyone at AWE gets all of it back as rebate or in some benefit.

    He also neglects the fact that if the result is cleaner air or roads less clogged, or fewer people being injured in road accidents which also impose upon the health system, and gets to avoid price shock inflation as a result of a sudden sharp and unforetold increase in oil prices or a fund that pays for the costs of sea level rise (moving ports and associated infrastructure) adjusting coastal development polices) then he gets other associated benefits. The tactic of the Liberals is to focus only on one part of the policy mix — the payment side of the ledger as if somehow, he is being asked to pour money down the toilet.

    Neil also avoids addressing the key issues about which there is no uncertainty at all. These are that:

    1. The Earth as a whole is warming at a rate unprecedented in at least 400,000 years and off a much warmer base than has been the case since at least the mid-Miocene c. 15-20 million years ago.
    2. There is no plausible way to account for this rapid warming that does not accord the dominant role to the uptick in CO2 and equivalent
    3. There is no plausible way to account for the uptick in CO2 from a steady 180-280ppmv during the last 800,000 years to the nearly 400ppmv we have now that does not include the combustion of fossil fuels by humans.
    4. There is no basis for thinking that if this upward treand is not halted, that the Earth will not keep warming, with all that implies for ecosystem services of value to humans. Indeed, we may be very confident that warming will accelerate for at least the next 100 years under the minimal action scenarios, due to positive feedbacks such as loss of albedo in the Arctic and Antarctic and the Alps, loss of permafrost and associated CH4/CO2 releases and a decline in the efficacy of existing marine and terrestrial carbon sinks, and through premature dieback of forested areas. In this scenario, it is very likely we will see a crash in biodiversity of fauna and flora, a loss in vegetation cover.

    This is the future about which Mr Fisher is indifferent and the legacy he is willing to leave to the next dozen generations to grapple with, not on the basis of any scientific expertise, for he admits he has none, but because he is bothered about the mere possibility of impositions from others on his lifestyle choices.

    Once upopn a time, when humans were truly primitives — perhaps 55,000 years ago before they devised their first fishhooks from bones joined with twine — Mr Fisher’s attitude might have been reasonable. The humans were few in number and the difference between life and death was surely very tiny. There was no government and no community in any sense we would find meaningful. There were no possessions to worry about and no legacy to conemtplate. One person’s choices made very little difference to anyone else and indeed, they lacked the language to even begin to manipulate each other, as far as we can tell.

    Mr Fisher’s politics and culture point back to this era of truly authentic individuals. This is, or ought to be, a neo-liberal wet dream — No taxes and no government, with each man armed to protect himself and his own pile of dung. It is truly sad that he cannot return there to enjoy its bounty in full.

  8. @Freelander

    For the record, I’ve no problem with scientific work being “audited” in a general sense. That’s what peer-review is about. It is what happens when someone tries to replicate another’s work, or apply an infrence to a suitable problem.

    So far, the cultural agnosophists have point blank refused to do this and still less have they been willing to publish work that would withstand such scrutiny. They have stood outside the gates and thrown rocks in the hope that someone will be hurt or the edifice will look a little shabbier to those who know nothing of engineering rather than build a more impressive rival city.

  9. @Freelander
    I suggest you have a look at, say, Science or Nature or any other reputable journal you care to examine. In it, you will find that there are regularly corrections, ammendments and clarifications complete with acknowledgements to those who found said errors – happens all the time. The errors may be minor and not affect the conclusions of the paper, or they may be so significant that they result in a retraction of the paper. In several cases, the so-called “deniers” have pointed out errors in published works that have been acknowledged by the author and no correction, ammendment or retraction (as appropriate) was issued in the journal where the original research was published or anywhere else. This is a distortion of the scientific record and is unacceptable in any field, regardless of it’s impact on public policy (which for our purposes here, we may currently ignore). One example of such is MBH98 – the “hockey stick” paper. Several people, including SM, have shown that the method used is a data mining technique, but that the statistics used do not take this into account; that one can feed trendless red noise into the algorithm and it will generate a trended output. If you care for arguement from authority, then Wegman has examined the issue and agrees with SM – impeach Wegman’s statistical knowledge at your own peril. Yet this paper has not been withdrawn, there has been no correction published, and the paper is still being cited to this day.
    Now, I realise that you may dispute that these facts are correct, but I would ask you for the moment to assume that it’s true (I assure you it is) – do you think this is in any way acceptable? To me, it is in no way acceptable from a scientific standpoint.
    From the CA blog, it’s apparent that someone is protecting these authors from normal scientific practice. An example: SM send such a notice to one of the leading journals in the “correspondence arising” sections. They sent it back several times asking to have it reduced in length. SM complied, adding a note to the effect that it was now extremely terse, dense and hard to follow. It was rejected for publication on the basis that – surprise, surprise – the reviewers found it difficult to follow and that it did not include sufficient references (such references having been removed at the request of the journal editor to keep the number of references down).
    Looks like a duck, walks like a duck, quacks like a duck, yet you apeear to be suggesting it’s a swan.

  10. “Impeach Wegman’s statistical knowledge at your own peril”

    Happy to do so. The social network stuff in the Wegman report was silly, and the whole genesis of the report (commissioned by far-right Senator Inhofe) suggests a consultancy with weak arguments for predetermined conclusions, very much at variance with those of the National Academy of Science assessment

    Note at Wegman has subsequently signed on to the “no warming since 1998” meme, which anyone who has any statistical understanding of time series can see is nonsense.

  11. I don’t have a problem with auditing. I think it is probably a good idea. However, there is a difference between being audited by someone who is simply trying to do a professional job and someone who is simply trying to engineer a ‘beat up’ of someone’s work.
    Without auditing, if they were serious, they could get the raw data from source do what they think is the best to get a final data set, and then present it to the world and see if they can get their work published in a good journal. Having done this, would they get a significantly different series? Assuming it was significantly different would their series stand up to scrutiny concerning the ‘adjustments’ they made. Given that the raw data would require aggregation and a variety of things before it would even give a series, there is no such thing as an unadjusted series. The raw observations would not give a series on the global temperature because there is no such thing as an instrument that can do a single reading and give global temperature. In science there is a good argument for more auditing, but the problem is that that would have to be paid for, and the people who can do it properly would find themselves better rewarded doing other things. Nevertheless, providing raw data to some data bank, as well as appropriate computer programs to take the raw data and transform it to its final form and to produce the various statistical results would be a good thing. But this is rarely done, so someone can’t be singled out for not doing it when almost no one at present does.

    I just had a read of a McIntyre paper criticizing the ‘hockey stick’ and it is very unclear what his point is. I suppose this means his paper is all wrong ‘because he hasn’t made it clear enough’. IMHO what was being done in the paper, by means of a criticism, was very strange indeed.

    The paper is at:

    Click to access mcintyre.grl.2005.pdf

    In this paper they discuss ‘red noise’ also called brownian noise, an example being a random walk. They seem to be suggesting that if the errors in the data are ‘red noise’ then at the end of the series the results could be mainly the noise (and the noise could wander up giving the hockey stick). This is true, to some extent. But why would the errors in the series be ‘red noise’ getting bigger and bigger by accumulation as time goes on? They don’t explain why and there doesn’t seem to be any non-magical why. A magical assumption after which, a magical result. But the result is only, if you make that magical assumption, the ‘hockey stick’ could be rubbish.

  12. @jquiggin

    Wegman is very qualified and should know his stuff. However, I suspect that he has some pretty strong philosophical views which came into play when he looked at this stuff. And that would be the reason he is out on his own. (I haven’t looked at his report, yet. Interesting that he is at George Mason.) Some very smart and able people can believe some very unbelievable things and say some unbelievable things when those things seem necessary to fit in with their personal philosophy.

  13. @Neil Fisher

    Having looked at one McIntrye paper I am not surprised at the reluctance of a journal like Nature to publish his correspondence. One trick that some academic do is write some rubbish comment to a good journal and if it is published the comment gets added to their CV as a publication. You can understand the reluctance of journals to reward this type of behaviour.

  14. @Fran Barlow

    1. The Earth as a whole is warming at a rate unprecedented in at least 400,000 years and off a much warmer base than has been the case since at least the mid-Miocene c. 15-20 million years ago.

    Anthropogenic CO2 was negligable in a climate sense until around 1940 or so. The change as measured by the instrumental record prior to this time (1900-1930 or there abouts) is statistically very similar to the period 1970-2000 (or there abouts). How this makes the latter “unprecedented” is therefore a mystery.

    2. There is no plausible way to account for this rapid warming that does not accord the dominant role to the uptick in CO2 and equivalent

    In order for this to be true, we would need a reliable record of GMST and CO2 that stretches back at least 1500 years, because there are natural variations in the proxy records that appear to cycle on time scales stretching from 30 years to 1500 years at least. There are no available proxy records that have the requisit temporal and spacial resolution and neither do they have have sufficient temperature resolution to negate the null hypothesis that this is natural variation. And without a sufficient degree of confidence in what is natural variation, there is no way to determine which part (if any) is anthropogenic.
    In any case, a 1% change in albedo would be sufficent to account for the measured change – and we simply do not have the data to know if this has occured.

    3. There is no plausible way to account for the uptick in CO2 from a steady 180-280ppmv during the last 800,000 years to the nearly 400ppmv we have now that does not include the combustion of fossil fuels by humans.

    There is no explaination for why around 60% of anthropogenic CO2 emissions have been absorbed, sequestered or whatever term you prefer and do not appear to be currently in the atmosphere. Nor is there any explaination for how this has been maintained for the period we have records for even though anthropogenic CO2 emissions have increased significantly over that period. With no explaination, there is no reason to assume that this will change in the future. With the top several km of the worlds oceans holding more CO2 than the entire atmosphere by around 3 orders of magnitude, even minor changes in solubility (a 0.1% change, for example, which would be extremely difficult to measure) would be sufficient to account for the changes.

    4. There is no basis for thinking that if this upward treand is not halted, that the Earth will not keep warming, with all that implies for ecosystem services of value to humans. Indeed, we may be very confident that warming will accelerate for at least the next 100 years under the minimal action scenarios, due to positive feedbacks such as loss of albedo in the Arctic and Antarctic and the Alps, loss of permafrost and associated CH4/CO2 releases and a decline in the efficacy of existing marine and terrestrial carbon sinks, and through premature dieback of forested areas. In this scenario, it is very likely we will see a crash in biodiversity of fauna and flora, a loss in vegetation cover.

    I see no reason to believe that we are outside the range of natural variability. I base this on historical, archeological and proxy evidence. But even if it was, and even should all the downsides you mention come to pass, there will undoubedly be upsides as well, such as greater crop poductivity. There is much mention of the bad things that will happen, but little (if any) mention of anything good, yet history and fossil records both tell us quite clearly that cold times are bad times for both nature and humans, while warm times are good for both. This is true for changes on the order of one order of magnitude greater than we have experienced during the entirety of the 20th C. (ie changes of around 8C)

  15. John, I just looked at your previous post which you linked to. So Wegman discovered that people who are leaders in their field write a lot of papers in their field and collaborate with other leaders in their field and, probably, are often used to refereeing work in the field. Who would have thought that?

  16. @jquiggin

    The social network stuff in the Wegman report was silly, and the whole genesis of the report (commissioned by far-right Senator Inhofe) suggests a consultancy with weak arguments for predetermined conclusions, very much at variance with those of the National Academy of Science assessment

    Then ignore the scoial network stuff, and focus on the technical statistical arguements.
    I find it interesting that you do not comment on who commissioned the NAS report and nor do you suggest that they might also have predetermined conclusions. I am not suggesting they did, but it is disingenious to suggest that one is tainted by such and not examine the possibilty that the other may be also.

    Note at Wegman has subsequently signed on to the “no warming since 1998? meme, which anyone who has any statistical understanding of time series can see is nonsense.

    Sorry John, but it most certainly is not nonsense. It may argueably be irrelevent in climate terms because of the short time scale, but from a purely statistical point of view, there is no statistically significant warming over this period. The data actually shows a slight downward trend, but it is not statistically significant. I believe that RealClimate itself showed this to be true, although I do not know who authored the post or when.

    @Freelander

    Wegman is very qualified and should know his stuff. However, I suspect that he has some pretty strong philosophical views which came into play when he looked at this stuff.

    This is the sort of thing I truely do not understand – if the data is equivicable enough that a minor variation in the analysts preconception can influence, by her choice of methodology, the outcome, then how can anyone suggest that their favoured outcome is “indisputable”, “clear as day” or whatever phrase you care to use? This is clearly a nonsense.

  17. @Freelander
    If you care to examine Wegman’s report yourself, you will see that he compares this to his own work (among others), which does not show similar patterns. While this does not prove anything due to the respective sizes and scopes of the fields involved, it certainly raises some concerns that bear investigation – which is what Wegman said.

  18. Oh, BTW all of you. You may have noticed that I am in the skeptical camp (may? he he!) and wonder why I am here. Well, I have noticed that this sort of forum is greatly enhanced by someone from the “other side” making the odd thoughtful post and steering well away from the ad-hom. I hope you will find I am performing this function here. I hope you will excuse the odd moment when frustration makes me stray into the ad-hom area or I get abusive – I try not to, but am only human. I trust our host will continue to allow me to post in good faith providing I stick to blog rules (which I admit I haven’t read, but I suspect are similar to most – don’t make accusation you can’t defend, don’t be abusive etc etc).
    Being in the minority here, I hope you will all appreciate that I am unable to repsond to each individual post directed at me, but I will do what I may as time allows. It is certainly more stimulating to be posting here than where I am part of the majority, and would thoroughly recommend such a course to anyone here who thinks they can take the heat. It ain’t easy, but it helps you keep an open mind and re-examine things you thought were settled.
    Cheers to all.

  19. @Neil Fisher

    Clearly, when the subject of climate change comes up, Wegman turns into a fruit loop. If you have a look at young earth believers and creationists you will find, among them, some very smart and able people who are able to do real science, mathematics and other things. That is, just so long as that science doesn’t move anywhere near their interpretation of the bible. Then, they suddenly become full blown fruit loops. Interesting phenomenon but there it is.

  20. To further respond on Wegman (and also on McKitrick, for example) when you see people making absurd errors which happen to produce conclusions congenial to them, it becomes pointless trying to check through everything they have written on the subject in question to try to see if some of it makes sense.

    I note further your suggestion that the US National Academy of Sciences, during the Bush Administration was subject to sinister influences of some kind. I’ve named names regarding Inhofe and Wegman. Would you care to do the same?

    As regards statistical significance, I’ve already dealt with that at length
    https://johnquiggin.com/index.php/archives/2010/03/05/list-of-the-clueless/

  21. @Freelander

    You can understand the reluctance of journals to reward this type of behaviour.

    Indeed I can. I would hardly consider SM’s behaviour to fit the mould you describe however.

  22. @jquiggin
    Err, forgive me if you find this offensive as it’s not meant to be, but this is an opinion piece and provides no evidence, no references and no data – in other words, it doesn’t seem to make any arguement other than “I’m right, you’re wrong, nah nah ne nah nah”. As I said, when even RC agrees that there is no statistically significant change and someone as well regarded in stats as Wegman is says the same thing, I find your arguement about as far from compelling as it is likely to get. Had you qualified your statement with “at climate time-scales” I would be happy to agree, however. Perhaps you feel that this is implied, in which case I would ask you to make it explicit. Then we can agree and drop the whole thing. If this is not the case, then we can perhaps agree to disagree if you do not wish to expand on your reasoning (or you could perhaps point me to a post that does contain evidence, references and data if you have already made one).

  23. @Neil Fisher

    If you take any data set and cut it down to a small enough set of observations, nothing is significant (statistically). But to talk about it not being (statistically) significant in that case as though you are saying something that is significant, is either a demonstration that you don’t know what you are talking about, or in Wegman’s case that you are either being dishonest or delusional.

  24. @jquiggin

    I note further your suggestion that the US National Academy of Sciences, during the Bush Administration was subject to sinister influences of some kind. I’ve named names regarding Inhofe and Wegman. Would you care to do the same?

    John, I specifically said that I was not making any such suggestion. Instead, I was commenting that it seems somewhat hypocritical to accuse one side of such tactics, while not mentioning even the posibility of the other side being inflicted with the same malady. It may not be, but you did not even mention it let alone provide information to refute any such claim in even general terms. I realise that you are probably not used to people being what may appear to you such a pendant, but in text based media such as this forum, it is very easy to convey the wrong impression, which is why I specifically stated that I was not suggesting what you subsequently stated I was. Please be more careful in the future.

  25. To put it another way those who say that the temperature change since 1998 has not been statistically significant are not saying anything meaningful.

  26. @Freelander

    If you take any data set and cut it down to a small enough set of observations, nothing is significant (statistically).

    Indeed – there are means to determine the confidence intervals one places on such time series trends though and I suspect that Wegman for one is well versed enough to understand this and apply it appropriately.

    But to talk about it not being (statistically) significant in that case as though you are saying something that is significant, is either a demonstration that you don’t know what you are talking about, or in Wegman’s case that you are either being dishonest or delusional.

    If it meets the statistical requirements for (in)significant, then in statistical terms it is (in)significant, regardless of how it fits into any larger collection of data. It is not dishonest or delusional to tell the truth. If this was the question asked, then he has answered honestly. If a different question, such as “how does this compare to the longer series?” was asked, he would have answered that honestly too, I am sure. Such context may indeed indicate that the noted trend may mislead the casual observer. Complaining that someone answered the question put to them because it disregards things that were not part of the question asked seems to indicate that it is you, sir, who have an agenda. If your arguement is that the question was inappropriate, then that is an entirely different matter and I would be happy to discuss it with you.

  27. @Freelander
    Indeed. I subscribe to the idea of “Trust, but verify”. This means: I don’t assume people are being dishonest or have an (political) agenda they wish to advance when they publish in the scientific journals, but I realise that people make mistakes (typos, transliterations, skipping lines – the list is endless) and so it is not inappropriate to have people check these things. Given the mistakes now coming to light in IPCC AR4, which was touted as the “most peer reviewed publication in history”, I think you can appreciate my point. When the stakes are high as they are for climate issues, the more people looking the better IMO. There is no dishonour in being wrong or making a mistake in science – as I said, it happens all the time, and finding these mistakes or errors is just as important to the advancement of science as the original contribution – sometimes more so.

  28. “If it meets the statistical requirements for (in)significant, then in statistical terms it is (in)significant, regardless of how it fits into any larger collection of data”

    This is a nonsensical claim. To take it to its logical extreme, if you tossed a coin 1000 times, and it came it came up heads every time, you could dismiss the conclusion that the coin was biased by observing that every single observation was statistically insignificant. What’s been done here is no better. Taking a time series that shows a statistically (and climatically) significant upward trend, those making this point observe that, if you break it up into short enough subsets, none of those subsets display a statistically significant trend.

    Honestly, you really need to study basic statistics before making claims about statistical significance. It’s a subtle concept, not at all simply related to “significance” in the ordinary sense of the term.

    The frequency with which this claim has been repeated illustrates an important point – most “sceptics” lack the basic training required to assess the evidence on these issues, and therefore choose a side on the basis of personal preference.

  29. @Neil Fisher

    But even if it was, and even should all the downsides you mention come to pass, there will undoubtedly be upsides as well, such as greater crop productivity. There is much mention of the bad things that will happen, but little (if any) mention of anything good, yet history and fossil records both tell us quite clearly that cold times are bad times for both nature and humans, while warm times are good for both. (typos corrected)

    Again, Free Air CO2 Enrichment (FACE) studies don’t bear this out. You simplify what is complex. It very much depends on which plants and which plant parts we are talking about before we can begin to model the likely scenarios for their hypothetical utility.

    Here is an entry level source that those interested in this question may consult,
    from the “climate crock of the week” series.

  30. @Neil Fisher

    Why is it difficult to accept post-hoc adjustments made to the base data, assuming they are as large as you claim?

    I do not have a difficulty with pos-hoc adjustments to the data – the difficulty I have is that such adjustments do not appear to be reflected in the uncertainty intervals when they clearly should.

    What makes you think they aren’t? Just because an adjustment is say, 1 deg C, does not mean that the uncertainty is increased by 1 deg C.

    And they are that large – a fact you can check for yourself at the NCDC, NOAA and GISS web sites.

    Those websites are fairly large and complex. Perhaps you could give a link to enable checking exactly what you’re talking about.

    data from the early 20th C can be volatile – it changes month to month as more data becomes available (it’s an artifact of the infill methodology, BTW). For proxy data, this would understandable, but for instrumental data?

    If you don’t like the data from the early 20th C then just use the data from 1964.

    Once again you misrepresent my point – which is that even if such adjustments are required and fully justified (IMO, they are not, but that is another arguement) I can see no reason why the adjusted temperature for 1930 (say) should be different in 1990 than it is in 2010.?

    Perhaps that says more about you than about the global average calculation methodology. How much do you know about calculating global average anomaly? And how much “correction” ar we talking about?

    By the way:

    GISS claimed error margin +-0.5C.

    You haven’t yet pointed out where this comes from. +-0.5C may apply to absolute global temperature, but anomaly is +-0.1C. Also:

    So what’s the issue with waiting 10 years if a century of warming is overwhelmed by such noise?

    As I pointed out, if climatically insignificant nose is filtered out by an appropriate low-pass filter, the noise comes nowhere near overwheling a century of warming. Your point is just another version of cherry-picking.

  31. @Neil Fisher

    There is no explaination for why around 60% of anthropogenic CO2 emissions have been absorbed, sequestered or whatever term you prefer and do not appear to be currently in the atmosphere.

    This is just bizarre. I’m astounded that anyone with a pretence of rationality could suggest there is no explanation for where most of the excess CO2 from anthropogenic emissions has ended up. Reading about ocean acidification should give anyone a hint. If you really believe there is no explanation then you’ve been taken in by nutcases.

  32. @jquiggin

    This is a nonsensical claim. To take it to its logical extreme, if you tossed a coin 1000 times, and it came it came up heads every time, you could dismiss the conclusion that the coin was biased by observing that every single observation was statistically insignificant.

    At 10 tosses, there is likely not enough data to constrain the confidence interval enough be be reasonably sure the coin is biased – I may just be (un)lucky. At 100 tosses, the confidence interval has narrowed and I have can have much better faith that this is indeed a biased coin. But what if the other 900 were tails? It would still be true to say that the 100 head tosses indicated the coin was biased, and we could place a confidence interval on that as well. If I asked you based on the 100 heads, would you say the coin appears to biased to heads? Or would you say that based on just the 100 tosses, it appears to be biased to heads, but taken in the context of all 1000 tosses, it appears to be biased towards tails? The first is answering the question as asked, the second may indeed be a more accurate description of the actual coin, but it is not an accurate answer to the question as asked. Pedantry? Perhaps, but not if you are asking a statistician!

  33. Are you a statistician? For the relevant purposes, I am (that is, I have research publications and teaching experience in econometrics), and these claims make no sense to me. More to the point, they represent familiar errors that every student is taught to avoid. Those promulgating the claims presumably know this, but correctly calculate that their audience does not.

    As regards your example, observing a sequence of 100 heads followed by 900 tails, I would conclude that the coin had been switched (or that some similar deception was being practised on me). The probability of such a sequence occurring from random tosses of a coin (with any kind of bias) is about the same as that of the molecules in my chair spontaneously reassembling themselves into a tree.

    We’ve spent a fair bit of time educating you on some basic points. At this point, if you’re not willing to concede that the “no significant warming” claim represents a basic misunderstanding of stats, I’m not willing to put in any more effort.

  34. @John Quiggin
    Let me try again. We’ll use temperature.
    Without going into the exact details, I’m sure we can agree that if we take a long enough look, we can probably create any trend we want by picking where we start. For instance, as we have been discussing, “no significant warming since 1998”. It’s true – the linear trend falls inside our confidence interval, so our premise is true. Do it with data back to 1900. Now the trend is different in sign and slope and is outside our condifence intervals so we have a significant trend. This does not mean that our first statement is false – it remains demonstrably true. I’m sure by examining the record, I could also find a starting point that is longer than 110 years and which had a negative slope, probably even more significant as well, than the data that shows a rise since 1900. This does not invalidate our second claim – that there is a significant trend in that data, yet you seem to be suggesting that it would!
    We could continue back and forth endlessly finding longer series and/or more significant trends until I eventually win because at the begining of the universe we have several billion degrees or more – foolish, I think you’d agree.
    So Wegman is not lying by supporting a claim that is demostrably true – it’s not his fault if you asked him the wrong question; it’s not his fault if the assumptions are incorrect because the assumptions were not his to begin with, but part of the question. He did some calculations and based on the data and assumptions supplied to him, answered accurately. What more would you ask of him? Because even had he in answered in nuanced form as you seem to believe he should have, do you really think that after being filtered through politicians and/or journalists, such nuance would have survived? “Yes, but…” becomes “Expert agrees…” quite easily.

  35. @Neil Fisher

    but taken in the context of all 1000 tosses, it appears to be biased towards tails?

    Pedantry? Perhaps, but not if you are asking a statistician!

    If you won’t be taught by Professor Quiggin you really ought to ask another statistician to correct your misconceptions for you Neil. If, that is, you can find the time in your busy pontificating schedule. What you’ve written on this thread, especially but not only your “interesting” opinion at #36, is just nonsense statistically speaking.

  36. @Alice

    And amusingly Alice, it is often the case that the deniers accuse us of wanting to return human life to the conditions of the pleistocene … when they fit that ambition so much better.

  37. @Fran Barlow
    Yes – they do. We seem to be attracting lots of amateur statisticians here making lots of errors who then want to deride those with real expertise. It must be turning the Profs hair grey. There seems to be a new disorder called education intolerance (like gluten intolerance).

  38. @Alice

    There seems to be a new disorder called education intolerance

    Dunning-Kruger Effect is the term du jour.

    The Dunning–Kruger effect is a cognitive bias in which “people reach erroneous conclusions and make unfortunate choices but their incompetence robs them of the metacognitive ability to realize it”. The unskilled therefore suffer from illusory superiority, rating their own ability as above average, much higher than in actuality; by contrast the highly skilled underrate their abilities, suffering from illusory inferiority. This leads to a perverse result where less competent people will rate their own ability higher than more competent people. It also explains why actual competence may weaken self-confidence because competent individuals falsely assume that others have an equivalent understanding. “Thus, the miscalibration of the incompetent stems from an error about the self, whereas the miscalibration of the highly competent stems from an error about others.”

    I like Betrand Russell here:

    The trouble with the world is that the stupid are cocksure and the intelligent are full of doubt

  39. @Fran Barlow

    Absolutely Fran……I love it. Ive seen it all here.

    “The trouble with the world is that the stupid are cocksure and the intelligent are full of doubt”

  40. Turning and turning in the widening gyre
    The falcon cannot hear the falconer;
    Things fall apart; the centre cannot hold;
    Mere anarchy is loosed upon the world,
    The blood-dimmed tide is loosed, and everywhere
    The ceremony of innocence is drowned;
    The best lack all conviction, while the worst
    Are full of passionate intensity.

    ……

    Yeats, THE SECOND COMING, 1919

  41. @frankis
    Frankis…You too are stiring up Irish passion with poetry. How easily we succumb… Two literatis with almost the the same point to make – how wonderful and interesting.
    Sure beats reading the indignantly indigenous here.

  42. @Alice

    Yes – they do. We seem to be attracting lots of amateur statisticians here making lots of errors who then want to deride those with real expertise.

    Experts disagree all the time. John appears to have a disagreement with Wegman. Fair enough. He told me that he’d shown Wegman wrong, yet the post he pointed me to did not contain any statistical arguement, it contained no description of an error Wegman had made, and it had no information about any faulty assumptions that were either supplied to or by Wegman as related to the question, and neither did it contain any logical arguement about why such conclusions, while accurate in technical terms, should be ignored. It was an opinion – a considered, learned opinion, but an opinion, not an actual rebuttal. You and he may find Wegman “obviously” wrong, but it is not at all obvious to me or anyone who reads the post John pointed me to as to why he finds it wrong, misleading, deluded or whatever other description anyone wants to put on it – the post simply does not contain that information. And I would remind you that John was replying to a comment I made to “impeach Wegman’s statistical knowledge at your peril” and pointing me at that post as such. If that post did contain an actual rebuttal, then I could ask another statistician to examine the details and give her opinion on the merits of the arguement, but I can’t do that because such rebuttal is not there. I cannot ask 1000, 100 or even 1 person to evaluate something that is not there, can I? No-one can evaluate an arguement that isn’t made. No-one – including John – has made a comment suggesting that I look elsewhere in this blog for more details on why it’s wrong; no-one has pointed out that such even exists here, let alone given a specific reference to where. I don’t expect that specific reference (though it would be nice), but before I start trolling through the archives, it would be nice to know that what I’m looking for actually exists – even a vague comment like “read the blog!” would indicate that the information is here, somewhere, but no-one has made such a comment. Perhaps it exists and perhaps it doesn’t – I don’t know. If it does, please say so and I will look for it – if you have a specific reference to where, that would be nice as it would save me considerable time, but even the general case will do if you can’t be bothered or don’t have the time.

  43. @Neil Fisher

    Wegman’s problem isn’t his statistical knowledge. His problem and the reason he is out on his own is that he simply is trying to use his authority, based on his knowledge, to mislead. Clearly he has reasons other than resulting from the rational use of his knowledge for believing and wanting what he believes to be true in this case. When he say the ‘no statistical difference’ meme, he knows that what he is saying is meaningless and grossly misleading to anyone who doesn’t really understand exactly what the statement means, which is most people. And when he propagates a statement like that he is simply being dishonest and is demonstrating his intention to mislead on this subject. As a non expert, trying to evaluate whether Wegman is right and everyone else is wrong on this topic, that he is knowingly seeking to grossly mislead on this subject should tell you something. But maybe it doesn’t?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s