This forum is about wrong numbers in science, politics and the media. It respects good science and good English.
The walls may be crumbling.
Can we survive their fall?
This is an alarming quote from the article attributed to someone or other:
“There is no cost to getting things wrong,” says Brian Nosek, a psychologist at the University of Virginia who has taken an interest in his discipline’s persistent errors. “The cost is not getting them published.”
It is this sort of thinking that has created the whole AGW scam. The missed point is that there seems to be more money in falsehood than truth.
The main theme of that article seems to be that peer review isn't sufficient and more replication studies are required in the practice of contemporary science. But as Bishop Hill pointed out a few weeks ago in responding to another Economist article (a leader article which is based on the article Brad has linked to), the Economist is a bit hypocritical in supporting public policy on subjects like AGW that are based on science that is pretty much only propped up by peer review.
Replication studies are very desirable, and it probably doesn't count as science if you can't replicate a claimed scientific work. However even replication is not good enough I think for public policy decision-making, because an erroneous assumption may be present in the original study and any replicated studies.
To give an example of how replication can be limited, consider the assertion that the Earth is (currently) about 6000 years old. This idea is based on a study in 1658 by a distinguished Biblical scholar, Archbishop Ussher of Armagh, that the Earth was created in 4004 BC (on Sunday, 23rd October at 9am to be precise). Some people may have heard of Ussher, but very few are probably aware that Ussher's study was independently replicated by others including Isaac Newton. Newton thought that Ussher's figure of 4004 BC was accurate to within a decade or so - "I do not pretend to be exact to a year: there may be errors of five or ten years, and sometimes twenty, but not much above". So it is possible to replicate a scientific-type claim, but the claim may still have a terrible validation status.
What a very interesting thread comment. Books have been written prior to the current AGW controversy about the failure of peer review in scientific activity. I don't recall the titles now but remember one book gave an account of the treatment of Jocelin Bell (incidentally, today, an AGW belief supporter) who discovered pulsars, but who's Nobel Prize was awarded to her supervisor, because he was the senior member of the team and she was just a subordinate doing her job.
In reply to Orde, my guess regarding the book you have read is that it sounds like a biography of Fred Hoyle. One of Hoyle's various hobbyhorses was that Jocelyn Bell was robbed of a Nobel prize in 1974, and the allegation got quite a bit of attention in the newspapers at the time. There was an ironic twist to that story, that people may be less aware of, in that Hoyle was himself mysteriously overlooked for the 1983 Nobel prize which was awarded to two of his former colleagues. The rumours were that Hoyle was snubbed by the Nobel prize committee because of the fuss he had raised a few years earlier over Jocelyn Bell, as mentioned in this article:
I'm pretty sure Fred Hoyle would not be keen on peer review, particularly the version of it that sprung up in the 1970s, as it would not be very helpful at all to someone with his creative maverick style.
Originally peer review was used (it was supposedly invented centuries ago by the Royal Society) mainly to filter out substandard papers, including possible hoax papers. But in the 1970s it picked up a new function - to reject dissident and speculative material from being published in the most well-read journals. The first victims of the new approach in the physics community were the relativity sceptics, who could get the odd paper discussing something like the twin paradox into the major physics journals in the 1960s, but then found similar papers were rejected in the early 1970s, forcing them to write books instead like Herbert Dingle did with his "Science at the crossroads". Hoyle wasn't so much of a dissident, his trademark was writing ideas man-type speculative stuff, and he would be able to get that type of material published fairly easily up to about 1970, but after that it became increasingly difficult.
In the 1990s peer review seems to have picked a further new function in that it is now regarded as a quality assurance-type activity which provides material worthy of adopting in public policy decision making. My guess at how this came about is that it is due to the modern day profusion of lobby groups. Decades ago there were far fewer lobby groups and they tended to stick to arguing from something like a moral standpoint, but in more recent years they've switched to making increasing use of technical arguments, bamboozling the political class with dubious facts and figures. The scientific establishment, noticing that it has now acquired some competition for providing scientific-type public policy advice from the lobby groups, has talked up the merits of its peer review activity. But though peer review is likely to result in more accurate and objective material than a report devised by some lobby group, it still only represents pretty superficial checking of the original work. The more detailed checking would be provided by replication studies, which the Economist points out are going into decline.