CERN of one of the major experiments looking for the Higgs Boson, sometimes referred to as the God Particle; an as-yet-undetected particle that is believed responsible for giving all other particles mass.
In the last several days rumors have swirled that CERN has the first conclusive evidence that the particle exists and that they have detected it at energies consistent with what most major theories have predicted.
But this is not the case. What we do know is that CERN has found what they are calling a hint of a detection.
The thing that you have to understand is that in the world of high energy physics (whether you're talking about high energy astrophysics, nuclear physics or particle physics) is that the data is accumulated in large quantities against an even larger background.
Sifting through the raw data and picking out data from the source (in high energy astrophysics this would be the gamma-ray or some other emission from a specific nebula, galactic nucleus or whatever; in the CERN experiment this would be the specific decay products that results from the Higgs) from the background is a lesson in complex statistical analysis.
One can never absolutely be certain that a signal is an actual effect, or a statistical fluctuation. But you can measure the likelihood that what you have is a detection.
Without getting into all the nitty gritty details of Li and Ma statistics, the ultimate point is that our discoveries are often expressed in gaussian sigma; a probability if you will that the detection is real, and not simply statistical fluctuations of the background signal.
For publication, most results need to have a confidence level above 5 sigma, meaning there is a about a 99.9999995% chance that the detection is real and not statistical. Sometimes �results above 3 sigma (99.7% confidence) are reported as an indication that there is an indication that the effect is real, but more study is warranted.
In this most recent CERN announcement two results were shown, but neither of them crested even the 3 sigma level. This is hardly convincing, and even CERN themselves are very quick to point out that they are not claiming a definitive detection, only that they have a hint that something is there.
More data and analysis is needed to determine if this is a real detection or not. Previously, a different result showed a detection at more than 4 sigma. But after further data was acquired and additional analyses were completed the effect was determined to be statistical.
This brings up another (and final) point. It is also important in these instances to complete multiple, independent analyses to verify the results. This such large amounts of data it isn't totally unexpected that a given analysis will produce some points that show significance above the 4 or 5 sigma level. But these artifacts can easily be eliminated with further analysis, whereas true detections will be verifiable with consistent results across all analyses.
So is this latest result exciting? Well, it is certainly encouraging, but shows that there is still a long road to go before convincing evidence is available. Stay tuned.
No comments:
Post a Comment