Talk:List of experimental errors and frauds in physics

remove

I removed Hafele-Keating_experiment from the list; the supposedly "not disputed" criticisms come from (as far as I can tell) two authors, one of whose sole publication on the topic was in the non-peer-reviewed popular magazine "Wireless World", the other of whose two non-peer-reviewed articles were in an obscure engineering journal and the also non-peer-reviewed "Physics Essays"; moreover, neither of the authors appear to agree with the thrust of the deleted entry, i.e. that H-K got the right answer via invalid techniques; rather, both authors appear to think relativity is wrong altogether. In any case, neither of these articles have any mainstream citations whatsoever, but are mentioned only on "Einstein-Was-Wrong" web pages and such. (This is in contrast to Eddington's irregularities, which are very well-known in the mainstream literature.) If anyone can find mainstream sources (WP:RS) for the idea that "H-K got their stats wrong", please cite it. Bm gub 19:32, 3 December 2007 (UTC)

My understanding was that Kelly's attitude was that the experimental outcome wasn't necessarily wrong, but that the experiment itself was unreliable. From memory, I think he said that the USNO guidelines issued the following year, if they'd been applied retrospectively to H-K, would have resulted in the clock that showed the greatest change being removed from the experimental analysis, and that without that single clock, most of the effect disappeared. Lots of people criticised Kelly's presumed motivation in disputing the validity of H-K, but I never saw anybody dispute his analysis of the subsequent USNO guideline, or dispute his analysis of how it would have affected H-K. ErkDemon (talk) 03:08, 27 June 2009 (UTC)

Actually, Eddington's experiment probably shouldn't be here either. It is understandable that it is, as there is "mainstream literature" severely criticising Eddington's experiment, but in recent years the pendulum has swung again and there are views now that Eddington has not been given due credit. What seems to have happened is that several myths sprung up about the experiment, one of which was picked up by the John Waller book "Fabulous Science" in 2002. For a full account of the myths surrounding the Eddington experiment, see Daniel Kennefick, "Not Only Because of Theory: Dyson, Eddington and the Competing Myths of the 1919 Eclipse Expedition," Proceedings of the 7th Conference on the History of General Relativity, Tenerife, 2005 (this article mentions the Waller book) I'm currently writing a Wikipedia article on the experiment (it's in my userspace at the moment), so hopefully that will help clear up some of the confusion surrounding this. Carcharoth (talk) 22:32, 20 February 2008 (UTC)

Totally agree. I removed Eddington. ScienceApologist (talk) 13:37, 7 April 2008 (UTC)

Eddington's result was problematic for a couple of reasons. The main one was that he had such a strong personal investment in proving Einstein's result that a sensible management decision would have been to put someone else in charge, if only to protect Eddington from the inevitable criticisms. If you have an experimenter who's that emotionally invested in getting a particular answer, then its not fair to them to put them in a position where they get to make judgement calls as to which data is included or excluded. It means that if they find that a dataset spoils their result, the temptation to exclude it may be extreme ... and even if they heroically manage to avoid temptation and get the result they wanted completely legitimately, others who aren't privy to all the data are likely to be (and entitled to be) sceptical. So the fact that he was allowed to be in that role was a weakness in the experiment's design - fairly or unfairly, it weakened the experiment's credibility.

There's also the issue about him supposedly neglecting to keep the plates that he decided not to use. On its own, that might not have been a huge issue, but since people were concerned over bias, and whether or not he might have exercised undue selectivity, it would have been nice if he could have produced the "bad" plates to demonstrate they were as bad as he said. Losing those plates was unfortunate.

And then of course, we have the huge experimental problems that the teams faced, and the lack of a decent baseline for measuring how accurate their photographs actually were. It's fair to say that daylight astronomy was not exactly a subject where people had a lot of experience, and one of the immediate problems that you have when you point a telescope at the skies in daylight is heating. If your optics warm and warp, then all the stars around the centre of the image will be displaced radially, yes? And then we also had the problem that we knew very little about the sun's atmosphere back then, and how far it extended. Solar atmospheric lensing, again, would displace the apparent positions of stars radially in the image.

So IMO Eddington's experiment fitted the original article title of "problematic physics experiments", and it was natural for it to be listed (even if the text then said nice things about the test). FWIW, I personally think that the new title "... experimental errors and frauds ..." is too specific, since it forces us to come down on one side or the other and excludes experiments that have been controversial in the past. "Problematic" was vague enough to be able to include things like the cold fusion experiments, and Newton's originally-disputed prism experiment, which were definitely problematic. ErkDemon (talk) 03:08, 27 June 2009 (UTC)

Cold fusion

There are several other cases listed on on Scientific_misconduct, notably Stanley Pons and Martin Fleischmann's cold fusion. —Preceding unsigned comment added by 69.3.190.146 (talk) 19:42, 27 July 2009 (UTC)

Polywater

Should Polywater be added to this list? —Preceding unsigned comment added by 184.100.219.37 (talk) 07:24, 7 February 2011 (UTC)

Interpretation of ice core data

Antarctic (Vostok station) ice core data was at one time quoted by several researchers and politicians to offer proof of a strong greenhouse effect due to carbon dioxide, and by inference, proof of dangerous anthropogenic global warming. It was later discovered that ancient carbon dioxide cycles lagged behind temperature changes and hence must have been a consequence, not a driver, of those temperature changes. In spite of the exposure of this error, we still see ice core data being quoted as proof of manmade global warming. Perhaps this should be included in the list of high-profile scientific errors, especially in view of its relevance to large-scale public spending today.

The original error (including the publicity material based on it by Al Gore) was probably a genuine oversight, but where this is still being quoted in spite of our new understanding of the data, it could be seen as deliberate fraud. --Anteaus (talk) 16:39, 1 June 2015 (UTC)

I don't know the ice core literature at all anymore. What's the most useful reference on this? Colin Rowat (talk) 20:28, 8 July 2015 (UTC)

CERN's 2012 speed of light results

Should the 2012 reports that experiments at CERN revealed faster than light particles be included, or were they resolved too quickly? Colin Rowat (talk) 20:30, 8 July 2015 (UTC)