Earlier this week, physicists at the Large Hadron Collider announced they’d found tantalising traces of a possible new fundamental particle — perhaps a heavier cousin of the Higgs boson, or the elusive graviton, a quantum carrier of the force of gravity.
The evidence comes from two separate, but complementary, experiments, known as CMS and ATLAS. Neither reported finding is solid enough to claim discovery, although the fact that both experiments see a slight hint of a particle in exactly the same spot in the data is promising. As Gizmodo’s Jamie Condliffe reported earlier:
The CMS team has prescribed the results a statistical significance of 2.6 sigma, while the ATLAS team’s has 3.6 sigma. These sigma values are a measure of how likely the results are to represent the presence of a hypothesized particle: for a result to be deemed a ‘discovery’ requires a sigma value of five, which equates to a 1-in-3.5-million chance of the result being a simple fluke rather an actual particle. A value of three is deemed to be an “interesting” result, with a greater chance of being a coincidence.
That said, there’s still a strong likelihood that this particular signal will disappear as the LHC scientists collect even more data. It happens all the time in particle physics — hence the cautionary note struck by pretty much every physicist who’s been interviewed about it not being time to pop the champagne corks just yet.
To get a better idea of why this happens, let’s take a closer look at how the LHC collects and analyses data. The machine collides protons at speeds close to the speed of light, and those high-energy collisions produce showers of particles. Physicists recognise the particles by the electronic signatures they leave behind, in the form of nuclear decay patterns. Quarks only exist for fractions of a second before they decay into other secondary particles. Since each quark has many different ways of decaying, there are several possible signatures, and each must be examined to determine which particles were present at the time of the collision.
A collision event with a photon pair observed by the CMS detector. Credit: CERN
That’s why the detectors used by ATLAS and CMS are needed to keep track of what’s happening and make sense of all the data. The detectors act as a filter, picking possible signatures of an unknown particle out of the tens of thousands of signals created every millionth of a second inside the accelerator. Physicists know precisely how many of each type of particle they should expect to see in the data; any excess above a certain threshold is a promising hint of possible new physics (like a new particle).
When all’s said and done, those signals show up as unexpected “bumps” in the data — that’s why experimental particle physicists often call themselves “bump hunters”. The thing is, it’s easy to see small “bumps” that aren’t really there; statistical artefacts crop up all the time, particularly during early data runs. The more data you have, the better the statistical analysis. If a small “bump” persists and gets bigger — the signal gets stronger — it’s much more likely that it’s the signature of a bona fide new particle.
Physicists talk about signal strength in terms of “sigmas”. As I wrote at Quanta for a 2013 article on the hunt for dark matter particles:
A signal’s strength is determined by the number of standard statistical deviations, or sigmas, from the expected background. This metric is often compared to a coin landing on heads several tosses in a row. A three-sigma result is a strong hint, equivalent to the coin landing on heads nine times in row. But many such signals weaken or vanish altogether as more data comes in and they turn out to be less statistically significant. The gold standard for discovery is a five-sigma result, roughly comparable to tossing 21 heads in a row.
Background noise makes the task more difficult. “A ‘signal’ is what you’re looking for. ‘Background’ is everything else that resembles your signal and makes it difficult for you to find it,” particle physicist Matthew Strassler wrote in a July 2011 blog post. Strassler’s favoured analogy is trying to find two friends wearing matching bright red jackets in a crowded room with a lot of other people also wearing bright red jackets. Sometimes you get random clusters of red jackets worn by strangers, causing you to mistakenly conclude that you’ve found your friends.
One of the ways physicists guard against these kinds of false positives is by employing what’s known as the “Look Elsewhere Effect”. You also have to calculate the probability that you would see something anywhere in the data — not just that you would see a bump in that particular location. Once that effect is factored in, the statistical significance of this latest possible signal drops to 1.2 sigma for CMS and 1.9 sigma for ATLAS.
The best you can say, in other words, is that the latest ATLAS and CMS results are inconclusive. At this point, it could go either way. For instance, earlier this year, there were reported hints of a possible heavier cousin to the Higgs boson, as well as glimpses of a possible supersymmetric particle dubbed “the edge”. Both those signals vanished in the latest analysis, after more data was added to the mix.
On the other hand, as Dennis Overbye observed in the New York Times, four years ago this week, both CMS and ATLAS reported tantalising hints of the as-yet-undiscovered Higgs boson, with “bumps” in the same signal range as the latest candidate. Six months later, they’d accumulated enough data to exceed the critical 5-sigma threshold and claim discovery.
We’ll probably know for sure by next summer, when both collaborations are expected to present results from the most recent data run at the LHC. “We’re in that interesting moment when all we can say is that there might be something real and new in this data, and we have to take it very seriously,” Strassler wrote about the new results. “We also have to take the statistical analyses of these bumps seriously, and they’re not as promising as these bumps look by eye.”
Or as Nima Arkani-Hamed, a theorist at the Institute for Advanced Study in Princeton, told the New York Times, “While we are nowhere near moving champagne even vaguely close to the fridge, it is intriguing.”