You might expect the Business Roundtable to be avid readers of that august weekly news magazine The Economist, and yet BR head honcho Roger Kerr was happy to write this in an op-ed published last month, apparently relying on British tabloid the Daily Mail as a source:
On top of all this is Climategate, which started with the leaked emails from the University of East Anglia’s Climatic Research Unit. Its suspended director Phil Jones has admitted that there has been no global warming in the past 15 years.
No he didn’t. Here’s The Economist on the subject:
Since I’ve advocated a more explicit use of the word “lie”, I’ll go ahead and follow my own advice: that Daily Mail headline is a lie. Phil Jones did not say there had been no global warming since 1995; he said the opposite. He said the world had been warming at 0.12°C per decade since 1995.
The Economist’s writer goes on to note that:
Anyone who has even a passing high-school familiarity with statistics should understand the difference between these two statements
One must presume, therefore, that Roger Kerr lacks that attribute, or is perhaps prepared to allow a good story to trump the facts. Not surprising when he lists in a Dominion Post opinion piece the experts the BR has brought to New Zealand to “balance” the debate:
Over the past 15 years the Business Roundtable has brought Richard Lindzen, Robert Balling, Patrick Michaels, David Henderson, Bjørn Lomborg and Nigel Lawson to New Zealand in an effort to inject some balance into the debate.
By their friends shall we know them.
Attention was drawn here to Kerr's repetition of the Mail lie a while back. The columnist "Curmudgeon" recently repeated it in the Dominion Post.
Haven't we been over this? No statistically significant warming since 1995. The fact that in a data set with lots of noise the last point is ahead of the first point does not prove a trend. The 0.12C per decade is simply this difference divided by the years. It seems it is the Economist writer who does not understand statistics. It seems warmists are clutching at straws.
(more to follow)
The significance test is to determine if noisy data does actually have a trend, or if the difference is caused by random movement. For example if you asked a random number generator to give you 15 numbers in an order you could claim a trend by looking at the difference between the first and last number (and dividing by 15). However a significance test should see through this mirage and prove the trend insignificant (unless the data set was not random but did involve a trend).
For people then to say, “he didn’t say no warming he only said no statistical warming” is fairly idiotic.
It’s a bit like if someone rolled a dice 15 times and the first role was a 2 and the last role was a 4 and they then turned around and said, “I’m getting better at rolling high numbers!”. Unless it can be shown that the trend exists beyond the realms of what’s randomly likely to happen without a trend, the trend doesn’t exist from a scientific point of view.
I don’t usually like referencing Wikipedia pages but perhaps some of you should read this:
http://en.wikipedia.org/wiki/Statistical_hypothes…
Yes but you can't rule out that there is no increasing trend (temp is completely random) either with this amount of data. So saying there is no warming since 1995 is also wrong
Tamino explains it well in this http://tamino.wordpress.com/2010/03/21/the-power-…
No, in statistics you accept the null hypothesis if you cannot reject it at X confidence level (usually 95%). If this attitude was not taken the required confidence level becomes 50%.
Saying there is no statistically significant warming between 1995 and present is correct.
But that's not what Kerr wrote: he claimed that Jones had "admitted that there has been no global warming in the past 15 years". And that is not true.
To write an entire blog over such a trivial difference seems nit picky to me. I guess he should have written "Jones admitted that he could not prove any global warming in the past 15 years, but I, Roger Kerr, would like to remind the reader that this does not mean that warming is defiantly not occuring at a level we can not prove, and it also does not prove the non-existence of lepracorns"
That is my point until there is a statistically significant trend up or down you shouldn't make definitive statements. Such as there has been no warming since 1995 or in this case misquoting someone who make a scientifically accurate statement by what I can only assume deliberately taking the comment out of context.
No, in statistics you accept the null hypothesis if you cannot reject it at X confidence level
No, you fail to reject the null hypothesis. That might be because you the null hypothesis is true or it might because the test/data you used doesn't have enough power to detect a trend.
You see for yourself. Open excel and type "=rand()" to one cell and ="rand() * 1.5" into the one beside it. Drag each formula down to fill 5 cells. Now, look up the formula for "TTEST" and compare the groups (you should choose one tailed test). You have just pulled samples from two populations, knowing one of them is 50% bigger than the other but you almost certainly didn't find a signficant difference. That's why you don't "accept the null"
I'm afraid the idiocy is yours, C3. The data since 1995 shows a warming trend. That trend (just) fails to meet 95% significance. The test in question is essentially asking "how likely is this trend to occur by chance". 95% means that 5 times out of a hundred it could occur by chance. Jones' calculation showed that the trend he found could have occurred by chance 6 or 7 times out of a hundred. The difference is immaterial. From 1994, the trend easily meets 95% significance.
All this proves is that in noisy time series you need a longish period of time to be sure of trends. Note that many of the people jumping on Jones statement are also keen to claim "cooling since 2002". The irony is obvious.
I will be posting more on trends in the global and NZ datasets soon, because the excellent Colin Sharples (aka CTG) has put together a new version of his superb trends applet (including showing whether trends are significant (95%) or not), and I'm going to give it its own page on the site. Watch this space.
C3PO is an exemplar of the saying attributed to Lincoln – "Better to remain silent and be thought a fool than to speak out and remove all doubt". It's time you went away C3PO, though of course the next robotic parlayer of denialist fantasy will no doubt get wheeled in to continue the "argument". My version of the "axiom of infinity" says that no matter how large the set of the deluded and ignorant is, another can always be appended to it.
OK so you are calling me a fool. Is there anything in particular that I said that you can point to and show it demonstrates foolishness? Or do you just disagree with me so you are calling me a name?
Re the Business Roundtable and Bjorn Lomborg, I recall that when Lomborg visited he said his usual thing: that money to reduce greenhouse gas emissions would be better spent on improving infrastructure and assisting economic development in 3rd world countries. The Business Roundtable published this with approval. So since then, I've been waiting for the Roger Kerr or one of his mates to advocate greater assistance to 3rd world countries. I'm still waiting.
PS: the usual WordPress.com login didn't work; I'll see about one of the alternatives. I'm the same Mark who has commented here previously.
Don't hold your breath…
Re log-ins. ID's web site says that the auto WordPress.com log-in feature is down at the moment, but you should still be able to log in manually using your WP.com log in/password.
This hair-splitting can only bring climate science into disrepute. There is either a trend or no trend. A trend is either significant or insignificant. If it is insignificant, then it is trivial, vexatious, time-wasting, and a mere distraction – better not mentioned.
A trend might conceivably be insignificant in statistical or scientific terms, even whilst it is regarded as significant in terms of philosophy or accounting or some other discipline.
Surely, on this blog, we are all speaking in scientific terms. In that context, an insignificant trend has no significance. It is immaterial and irrelevant. End of story.
You point to the different meanings of significance in this context, and then continue to confuse them! Far out, as the young folk once said…
There is a warming trend in the data. Take it over 16 years or longer and it easily passes this particular test of significance. Failing the test does not mean that there's no trend.
"Failing the test does not mean that there's no trend."
No of course not. That is not what any of us said. We said it means there is no significant trend.
Failing the test, as you have rightly said, means that trend cannot be ruled out as occurring randomly and independent of the variable being studied (time in this case).
So although the existence of pixies cannot be proven to be fictional, it can also not be proven to be non fictional.
OK so you recognise that there is inusfficient data to determine a trend at the 95% level with 15 years of data.
Using 1995 as a starting date I am willing to bet $100 that when a statistically significant (95%) emerges from the global temp record (anyone of the 4 I even include UAH) it will be a warming trend.
Doug
Doug
Hmmm not really how scientific hypothesis testing works……..
Perhaps but it demonstrates 'moderate' courage of convictions.
Forget pixies C3, try understanding statistics first:
http://tamino.wordpress.com/2009/12/15/how-long/
And some context for lurkers:
http://www.skepticalscience.com/Phil-Jones-says-n…
The contradictory claims of deniosaurs are absurd, one minute they claim Hadley Centre is a bunch of frauds with faulty data, the next it's okay to use the same data if they think it supports their wishful thinking. Try 15 years with GISTEMP.
So are we going to get this garbage from Roger every year?
Next year (2011): – there has been no significant warming since 1996.
In 2012: there has been no significant warming since 1997.
etc
Surely even C3P0 will see the achilles heel of the statement by 2013.
I guess someone fixed the commenter access routine because this went straight in for me [ I hope 🙂 ]
Now I hope you folks don’t mind my asking this question, but in regard to the lie-telling trend known to flash around from time to time, does this business roundtable character have a track record in this department.
I am of course referring to beyond science, which would appear to be a more appropriate realm for such activities. Which is not to excuse a fellow’s ignorance.. yet which invites his admission to it.. ignorance that is.
For the last 15 years the Roundtable (and that means Kerr) have been arguing a sceptic line — see the list of people they've invited to NZ. Kerr also has plenty of "previous": see this post for rubbish in a submission to Parliament. It includes one of my favourite lines:
Richard Lindzen, Robert Balling, Patrick Michaels, David Henderson, Bjørn Lomborg and Nigel Lawson
It's an orchestrated litany of…………denial
Gareth,
Tks for that observation. I wasn’t overlooking the names you mentioned, deeming them more ‘noise’ than anything. Noise vulgare one might say.
The lack of understanding of exactly what the Copenhagen Consensus was in that quote – heck and as part a NZ Parliamentary submission! – speaks volumes. What do the R/T members here think about that!
At my kindest the very best one could rate the quote’s import would be that it focuses as a politics driver. Which kind of puts the fellow well and truly into the strictly politics box. And the realm of lies.. as and when needed. Perhaps adding another frame to your blog’s call.
Politics can go so well with unscience, a term borrowed from the CURRA launch several days ago.
ps: Fifteen years..? Would that by any chance happen to coincide with statistical significance of lie-telling trend.?