View of the Tevatron at Fermilab. (Image courtesy of Fermilab)
By James Dacey
With the sheer scale of its machinery and its extensive international collaborations, accelerator physics is now a highly visible part of mainstream science. A recent episode at Fermilab reminded me, however, that the scientific results could never be as clear-cut as the facilities that produced them. I couldn’t help but feel a little bit frustrated by this, but perhaps that says more about my short concentration span.
Earlier this week, we reported new findings from Fermilab’s D0 experiment, which claimed to have gathered the strongest evidence yet for CP violation beyond the Standard Model. CP violation helps to explain the fundamental difference between the behaviour of a particle and its antiparticle. It explains why matter survived in the universe after the Big Bang, when matter and antimatter were created in equal amounts and should have annihilated completely.
So as you can see the implications of this research were fairly humongous. Hence physicsworld.com and a whole load of other sites covered the story.
However, last Monday – without us realizing – a rival experiment at the Tevatron had already poured cold water on the D0 celebrations by saying that the results were void. Let me explain.
D0 had looked for asymmetry in the production of muons from the decay of B mesons and anti B mesons, and they reported a CP violation that was 3.2 standard deviations larger than what is predicted by the Standard Model.
The particle physics community got very excited because their results had included decent measurements, for the first time, of the decay of Bs mesons, which theorists have long touted as an excellent place to look for CP violation. It seemed that the physics was finally moving beyond the Standard Model.
But before D0 had time to wind down their celebrations, Gavril Girgiu of the CDF experiment was in Turin addressing a particle physics conference about the CDF analysis of the same meson decays. The difference was that CDF had seen nothing out of the ordinary in their results, and their sample size was twice as large – from 5.2 femtobarn of Bs decays, they record CP violation that is within 0.8 standard deviations of the Standard Model.
Now, don’t get me wrong here. I fully realize that experimental physics can only ever move forwards by the proposition of new phenomena followed by its confirmation by other experiments. But, given the expectation that surrounds particle physics – fuelled largely by the high profile of its facilities – I just couldn’t help but feel a bit frustrated by these events. It was just a reminder that despite all the exciting questions and mind-blowing implications of particle physics, the real science so often boils down to more mundane concepts like statistical significance.
Perhaps there is a better way for the particle physics community to communicate that this part of the science can be fun too?
The danger is that people like me, who are intrigued by this weird and fascinating area of science, have come to expect every new result to be as certain as the facilities that have produced them.
Hi, i think the unit of sample size, “twice as large –
from 5.2 femtobarn” is inverse femtobarn.
This reminds me of the recent dark-matter dispute reported at http://physicsworld.com/cws/article/news/42554 , and is perhaps an example of a general issue. People are hungry for breakthroughs, and others need to be able to demonstrate some kind of result. That’s life I suppose, it’s an imperfect world, people are people, and physicists are too.
To correct some misunderstandings in this article: The CDF experiment can not perform the same measurement because they cannot invert their magnetic field. What CDF measures is a another quantity which might be related to what D0 has seen. The size of the D0 sample was 6.1 inverse femtobarns and the CDF measurement uses 5.2 inverse femtobarns, so the sample is not ‘twice as large’. It is actually smaller. Furthermore, the new CDF measurement is fully consistent with the D0 measurements. So, this CDF measurement, though interesting, has not added much more information, and it certainly cannot decide whether the D0 result will hold or not. The problem is not how physicists communicate their science, the problem is that writers like controversies, and they try to create them even if there are none. And yes, statistics is important and it has been made very clear by the D0 physicists that this needs to be confirmed with more data and by other measurements. But this is the excitement of science, just the opposite to what the writer seems to think, and that’s how science has progressed for hundreds of years.
Can you give any links to the “true story”, particlephysicist? I dig up things like the link below and it’s more of the same:
http://resonaances.blogspot.com/2010/05/cdf-says-calm-down-everybody.html
The real message from the ongoing D0/CDF controversy is that we need two independent experiments pursuing the same physics objectives but with different measurement and analysis techniques. This internal check is the best measure against getting ahead of oneself. If an experimental result can not be confirmed by other independent experiments, then it an artefact
and not real physics. I just want to remind what Val Telegdi once quipped (by the way, also in the context of CP violation): There are not only superweak theories, there are also super weak experiments!
In my opinion there is not enough experimental data to prove or disprove the claims. The confidence levels cited in the article are on that one set of experimental data – this only judges the validity of one experimenters results. D0 and CDF (although differing experiments) are just two sets of data. What is needed is more data from the LHC.