2 ms·
These do no represent a "conspiracy" to wrongly withhold data, because it does not represent: 1. Wrongly withholding data. They deleted data of questionable va
by KirinDave 17y ago
These do no represent a "conspiracy" to wrongly withhold data, because it does not represent:
1. Wrongly withholding data. They deleted data of questionable value that they knew could and would be misinterpreted by the disreputable people that, despite repeated attempts to reconcile with, had remained stubbornly antagonistic.
2. A conspiracy. It's only a conspiracy if it's kept secret, and as others have pointed out in this thread, the declines they corrected for were publicly acknowledged in the CRU's works. In order for there to be a cover-up operation and a conspiracy... there needs to be an actual cover-up.
If you see more evidence of a real conspiracy when you read about the loss of the "original" (although I think the proper word is "raw") data, I'd like to offer an alternative perspective. Having worked on climate-anlysis and recording systems for Lockheed Martin (in 2003-4 for the RSA2 & AWIPS contracts) I can tell you that it's not unusual to throw the raw data out. That data is noisy, often produced by poor equipment, and frequently the systems used to gather it have "curious" false reporting characteristics. We only kept interpolated data as well, as that was the only data of any value. This is in sharp contrast to the financial data you like to compare it to: financial data doesn't have significant noise and every transaction is of clear value independent of the whole. That is not what climate data is, and continuing comparisons along these lines suggest deliberate ignorance.
Most scientists simply don't see any value in this obviously noisy raw data. They're not used to having a bunch of belligerent charlatans try and anomaly hunt their data for even slight inconsistencies and then be told their life's work is meaningless and compromised. The way they react to this once they come into contact with special interest groups trying to fight science varies. Some go very aggressive (like Myers and Dawkins), some simply leave the field, some do their best to ignore the problem.
- yummyfajitas 17y ago1. Since they published papers that the IPCC used based on the raw data, it's apparently very valuable. Further, other scientists might want to try different types of analysis on the same raw data. They might get better or different results. You are simply wrong that raw data is not valuable. 2. These were private emails between scientists at the CRU regarding the deletion/withholding of data. They signaled an intent to delete data rather than give it to people skeptical of their work, and intent to evade their FOIA obligations. If you really dislike the comparison to financial companies, lets compare to Big Pharma. The data is messy, and they also have belligerent charlatans (aka lawyers) trying to anomaly hunt for a big financial reward. Would you accept this: "I think I'll delete the file rather than send it to the FDA". Fast forward a little bit, and the data has been "lost". "What little malice there is in the equation is a bunch of scientists frustrated with silly FDA rules designed to avoid any bad publicity"?
- KirinDave 17y agoYour response to #1 is nonsensical. Please reconsider your response. Also, could we please stop working in flawed analogies? You've tried to take my statement and substitute in things you consider similar to make a point. But the problem is, they're not similar at all except that there are people called "Scientists" involved. Different fields demand different levels of transparency, archival, and data fidelity. And the decision to throw away noisy or corrupted data is not unusual even in the medical profession. Go ask St. Jude CRMD for their raw electrocardiogram data for every pacemaker and implantable defibrillator design run. Even the FDA doesn't demand it, and I know for a fact they delete the majority of the raw analogue data. It is not important. We can reconstruct the data from previously published data and code, if you want it so badly. The operations done in the code are not particularly difficult to reverse to equivalent data. Even then, would that constitute evidence to you? What's to stop a conspiracy driven group of scientists from simply fabricating realistic-looking raw data? This is just a talking point, and everyone who's not out here with an axe to grind knows it.
- yummyfajitas 17y agoYou've tried to take my statement and substitute in things you consider similar to make a point... Different fields demand different levels of transparency... Could you explain why you feel climate science deserves low levels of transparency, and why conspiracies to delete/withhold data here are acceptable? Regarding medical trials, they do not delete the data they process. A concrete example: consider an fMRI study. The MRI spits out analogue k-space data. The MRI converts this to x-space data (pictures) which are then processed by whoever is running the study. It's legit to dump the k-space data (it's not part of his input), but he can't dump the x-space data. In contrast, I work directly on the k-space data, so I keep it. The rule of thumb is "keep your direct inputs". This is why I say the data was important - it enabled Mann/Jones/etc to publish important results. When (part of) it was made available to Steve McIntyre, it also allowed him to correct mistakes in the results of Mann. Even then, would that constitute evidence to you? What's to stop a conspiracy driven group of scientists from simply fabricating realistic-looking raw data? Unfortunately, there is very little which stops fraudulent scientists. Peer review is not really set up to do this. The best check is openness (release of code and data) and independent attempts to replicate published results. Unfortunately, the folks at CRU conspired to prevent this from occurring. Econometrics journals seem to be taking good steps in the right direction: http://www.econometricsociety.org/submissions.asp#Replication http://www.econometricsociety.org/submissions.asp#Replicatio...