6 ms·

You obviously didn't read the next paragraph: "Writing computer code to process the data "took less than two days and produced results similar to other indepen

You obviously didn't read the next paragraph: "Writing computer code to process the data "took less than two days and produced results similar to other independent analyses. No information from CRU was needed to do this"." Personally I can understand them not wanting to "fully comply", given the number of FoI requests they would've received and their tone. Have you seen the vitriol over at Climate Audit? edit: It gets even better a bit further down: "Asked whether it would be reasonable to conclude that anyone claiming instrumental records were unavailable or vital code missing was incompetent, another panel member, Professor Peter Clarke from Edinburgh University, said: "It's very clear that anyone who'd be competent enough to analyse the data would know where to find it. It's also clear that anyone competent could perform their own analysis without let or hindrance.""

underdown 16y ago
You mean like how research 2000 hasn't complied for kos? Its not science if its not open and independently verifiable. Raw data doesn't a computer model make.
roboneal 16y ago
The number and "tone" of the requests is irrelevant. If they want to ignore or not "fully comply" with FOI requests - they shouldn't take public funds.
- tome 16y ago
  If you're a university researcher in the UK it's impossible not to take public funds. All universities (bar one) are publically funded.
  - yummyfajitas 16y ago
    Then don't be a university researcher. In fact, if you don't want to reveal your top secret source code/data, don't publish in science journals. It's utterly ridiculous that journals still exist which don't require source/data to be released.
    - ant5 16y ago
      The data was available. If you want to independently verify it, the best way is to write your own code and see if it spits out the same/similar results.
  - roboneal 16y ago
    "All universities (bar one) are publically funded." That's a problem isn't it. Maybe the UK should fix that.
- moxiemk1 16y ago
  If only that logic worked on the DHS...
- ant5 16y ago
  In theory, yes. In practice, people are only human, and tend to not react well when faced with a vitriolic "tone". The police may be required to protect and serve, but try berating a police officer and see how well he serves you when your house gets robbed. He shouldn't treat you any differently, but he probably will.
- MichaelSalib 16y ago
  So, if each climate scientist had to field 5000 FOI requests per year, that would be no problem, right? Are you saying it would serve the public interests to spend money paying scientists to research while also requiring them to spend 99% of their time responding to FOI requests? It seems to me that we all agree that the number and tone of requests is relevant. I think that everyone here will accept that once the number of requests prevents scientists from doing any research, than we have a serious problem. So the question is not "is the number or tone relevant?" but "what fraction of scientist time should we expect researchers to spend responding to FOI requests before we consider them unduly burdensome?"; given this question, tone seems very relevant. If researchers are responding to questions from professionals who have a good understanding of the fundamentals and are scientifically literate, then we should expect them to be more willing to respond to inquiries. But if they're responding to scientifically illiterate people who don't understand the basics of climate modelling, then I don't think we should expect them spend any time at all. I mean, the FOI process can be abused. I think it is important that we differentiate between using the FOI process as a way of harassing researchers and serious questions from people who understand enough to actually deal with the data. In that sense, requiring people to write their own code (which is better for finding errors anyway) serves as a way to tax the harassers without unduly burdening real scientists. Real scientists will probably have planned on writing their own code anyway, but people abusing the FOI process for harassment probably don't have the capability to do so. In my own neighborhood, I've watched crazy citizens groups use demands for data as a way to harass experts. They go to town meetings and yell and scream about data. When experts give them the answers they need (which costs time and money), they start poring through the data looking for inconsistencies. Inevitably they find something that looks suspicious if you're completely ignorant of math and science and then they pounce on it at the next town meeting. "Gotcha!" they say. But they're ignorant people who don't understand basic math and they inevitably turn out to be completely wrong. Most of the time, the truth is exactly opposite what they've claimed. Meanwhile, an expert has to explain in great detail why exactly they're wrong. The whole thing is just a huge waste of time and taxpayer money. It doesn't cost them anything to demand data and when any expert fails to give them everything (including the kitchen sink) because they're worried that these ignorant people will misunderstand things, they only scream all the louder. Public policy is hard. Even very simple questions like "will this development make traffic on this side street worse" is actually fairly tough to answer. The vast majority of people don't have anything close to the skillset needed to intelligently review an analysis of these questions. But many think they do.
yummyfajitas 16y ago
No, you obviously didn't read the paragraph you quoted. It suggests that in addition to withholding data, they also withheld source code. The fact that maybe you could recreate the code or download data from elsewhere does not change the fact that the CRU did withhold both source code and data. As for the "number of FoI requests they would've received and their tone", that sounds like something apache can handle: GET /public_data/hockey_stick.tar.gz HTTP/1.1 Host: www.cru.uea.ac.uk/ Tone: HIGHLY VITRIOLIC
- ant5 16y ago
  If the point is to independently verify the results, you should write your own code to grind it and compare it to CRU's results. anthonyb already quoted the relevant section: Writing computer code to process the data "took less than two days and produced results similar to other independent analyses. No information from CRU was needed to do this". If a qualified researcher can reproduce the results in two days, I'd suggest that anyone who can't is not sufficiently qualified to be evaluating the research's efficacy in the first place.
  - yummyfajitas 16y ago
    I disagree that source code release should not be required, but that's a side issue. The main point I'm making here is this: the title of this article (and the report) is just not true. The CRU folks withhold source code and data, and the report even admits this. Not only did they not respond simply because of the "number" or "tone" of requests, but they actively avoided releasing data. From one of their emails: "And don't leave stuff lying around on ftp sites - you never know who is trawling them [...] If they ever hear there is a Freedom of Information Act now in the UK, I think I'll delete the file rather than send to anyone."
    - ant5 16y ago
      If the data you need to reproduce their results is publicly available, then it isn't withheld, is it? The purpose of the three separate inquiries was to separate fact from out-of-context quotes like the ones you're still trying to leverage. I think the results are clear. The researchers should have been more open, but they didn't do anything unethical and anyone qualified could have reproduced their results using widely publicly available data.
motters 16y ago
Indeed. And I did exactly as this report suggests and carried out my own independent analysis, regardless of whether CRU did or didn't make their data/software available. http://sluggish.homelinux.net/index.php?title=Tempgraph http://sluggish.homelinux.net/index.php?title=Tempgraph The station data and temperature data were already out there in the public domain.
- roboneal 16y ago
  Is it safe to assume from your resulting graph that "warming" in England has been fairly steady from 1750 and the mean temperature actually rose more dramatically from 1750-1850 than in modern times?
  - motters 16y ago
    Not really. Because of the high variance it's difficult to make claims of that sort. You have to bear in mind that prior to the late 19th century Stevenson screens were not in use, and thermometers themselves were less accurate. Also it's pretty clear that there are systematic biases in the data, meaning that WMO stations on average move closer to the equator over time. I havn't checked whether this bias exists for the UK data, but it does for the overall data set.

underdown 16y ago

You mean like how research 2000 hasn't complied for kos? Its not science if its not open and independently verifiable. Raw data doesn't a computer model make.

roboneal 16y ago

The number and "tone" of the requests is irrelevant. If they want to ignore or not "fully comply" with FOI requests - they shouldn't take public funds.

tome 16y ago
If you're a university researcher in the UK it's impossible not to take public funds. All universities (bar one) are publically funded.
- yummyfajitas 16y ago
  Then don't be a university researcher. In fact, if you don't want to reveal your top secret source code/data, don't publish in science journals. It's utterly ridiculous that journals still exist which don't require source/data to be released.
  - ant5 16y ago
    The data was available. If you want to independently verify it, the best way is to write your own code and see if it spits out the same/similar results.
- roboneal 16y ago
  "All universities (bar one) are publically funded." That's a problem isn't it. Maybe the UK should fix that.
moxiemk1 16y ago
If only that logic worked on the DHS...
ant5 16y ago
In theory, yes. In practice, people are only human, and tend to not react well when faced with a vitriolic "tone". The police may be required to protect and serve, but try berating a police officer and see how well he serves you when your house gets robbed. He shouldn't treat you any differently, but he probably will.
MichaelSalib 16y ago
So, if each climate scientist had to field 5000 FOI requests per year, that would be no problem, right? Are you saying it would serve the public interests to spend money paying scientists to research while also requiring them to spend 99% of their time responding to FOI requests? It seems to me that we all agree that the number and tone of requests is relevant. I think that everyone here will accept that once the number of requests prevents scientists from doing any research, than we have a serious problem. So the question is not "is the number or tone relevant?" but "what fraction of scientist time should we expect researchers to spend responding to FOI requests before we consider them unduly burdensome?"; given this question, tone seems very relevant. If researchers are responding to questions from professionals who have a good understanding of the fundamentals and are scientifically literate, then we should expect them to be more willing to respond to inquiries. But if they're responding to scientifically illiterate people who don't understand the basics of climate modelling, then I don't think we should expect them spend any time at all. I mean, the FOI process can be abused. I think it is important that we differentiate between using the FOI process as a way of harassing researchers and serious questions from people who understand enough to actually deal with the data. In that sense, requiring people to write their own code (which is better for finding errors anyway) serves as a way to tax the harassers without unduly burdening real scientists. Real scientists will probably have planned on writing their own code anyway, but people abusing the FOI process for harassment probably don't have the capability to do so. In my own neighborhood, I've watched crazy citizens groups use demands for data as a way to harass experts. They go to town meetings and yell and scream about data. When experts give them the answers they need (which costs time and money), they start poring through the data looking for inconsistencies. Inevitably they find something that looks suspicious if you're completely ignorant of math and science and then they pounce on it at the next town meeting. "Gotcha!" they say. But they're ignorant people who don't understand basic math and they inevitably turn out to be completely wrong. Most of the time, the truth is exactly opposite what they've claimed. Meanwhile, an expert has to explain in great detail why exactly they're wrong. The whole thing is just a huge waste of time and taxpayer money. It doesn't cost them anything to demand data and when any expert fails to give them everything (including the kitchen sink) because they're worried that these ignorant people will misunderstand things, they only scream all the louder. Public policy is hard. Even very simple questions like "will this development make traffic on this side street worse" is actually fairly tough to answer. The vast majority of people don't have anything close to the skillset needed to intelligently review an analysis of these questions. But many think they do.

yummyfajitas 16y ago

No, you obviously didn't read the paragraph you quoted. It suggests that in addition to withholding data, they also withheld source code. The fact that maybe you could recreate the code or download data from elsewhere does not change the fact that the CRU did withhold both source code and data. As for the "number of FoI requests they would've received and their tone", that sounds like something apache can handle: GET /public_data/hockey_stick.tar.gz HTTP/1.1 Host: www.cru.uea.ac.uk/ Tone: HIGHLY VITRIOLIC

ant5 16y ago
If the point is to independently verify the results, you should write your own code to grind it and compare it to CRU's results. anthonyb already quoted the relevant section: Writing computer code to process the data "took less than two days and produced results similar to other independent analyses. No information from CRU was needed to do this". If a qualified researcher can reproduce the results in two days, I'd suggest that anyone who can't is not sufficiently qualified to be evaluating the research's efficacy in the first place.
- yummyfajitas 16y ago
  I disagree that source code release should not be required, but that's a side issue. The main point I'm making here is this: the title of this article (and the report) is just not true. The CRU folks withhold source code and data, and the report even admits this. Not only did they not respond simply because of the "number" or "tone" of requests, but they actively avoided releasing data. From one of their emails: "And don't leave stuff lying around on ftp sites - you never know who is trawling them [...] If they ever hear there is a Freedom of Information Act now in the UK, I think I'll delete the file rather than send to anyone."
  - ant5 16y ago
    If the data you need to reproduce their results is publicly available, then it isn't withheld, is it? The purpose of the three separate inquiries was to separate fact from out-of-context quotes like the ones you're still trying to leverage. I think the results are clear. The researchers should have been more open, but they didn't do anything unethical and anyone qualified could have reproduced their results using widely publicly available data.

motters 16y ago

Indeed. And I did exactly as this report suggests and carried out my own independent analysis, regardless of whether CRU did or didn't make their data/software available. http://sluggish.homelinux.net/index.php?title=Tempgraph http://sluggish.homelinux.net/index.php?title=Tempgraph The station data and temperature data were already out there in the public domain.

roboneal 16y ago
Is it safe to assume from your resulting graph that "warming" in England has been fairly steady from 1750 and the mean temperature actually rose more dramatically from 1750-1850 than in modern times?
- motters 16y ago
  Not really. Because of the high variance it's difficult to make claims of that sort. You have to bear in mind that prior to the late 19th century Stevenson screens were not in use, and thermometers themselves were less accurate. Also it's pretty clear that there are systematic biases in the data, meaning that WMO stations on average move closer to the equator over time. I havn't checked whether this bias exists for the UK data, but it does for the overall data set.