Hi Cheryl, this is my understanding of how this works:
The Higgs is one of those particles that can't be detected directly because it will not tend to survive long enough to hit a detector. This is okay because we can calculate, when the particle decays, what it decays
into. So if you look in the right places, you will find descriptions of the various paths for "production" of various kinds of particles. The idea is that at a certain energy scale there are a certain number of ways a Higgs could come into being, and a certain number of things that are likely to happen when a Higgs is produced. Since we know a lot about the Higgs, we can calculate ahead of time what those things will be.
I can't find right now a description of the paths we'll likely see at the LHC, but
here's a description of how they're looking for the Higgs at the Tevatron, from the blog of a scientist there, which should give you a rough idea.
The short version of the Tevatron description as I'm reading it is: the Tevatron particles crash, and the energy of that crash could produce a number of things. Among the things it could produce is a Higgs, or it could also produce a W Boson and a Higgs, or maybe a Z Boson and also a Higgs. Meanwhile, once the Higgs comes into being, it will last for a certain amount of time, after which it could decay into a b-quark and a b-anti-quark, or it could decay into a pair of W bosons. Of course, the W bosons and such aren't directly observable either! They decay into other things... some of which decay into other things... eventually, all this decaying is done, and the particles that are left over are long-lived ("long-lived" meaning "long enough to travel a a few feet away to the detector") things like neutrinos. THESE are the things that the detector detects!
So basically, you're running this detector. With each collision you get a weird smattering of particles hitting the detector, and for each particle your detector registers things like its energy, its angle, whatever. And you sit down with a mathematical model that has a long, long list of all the different things that could possibly be produced in a collision; and for each of those things that could be produced, it has a list of "decay channels" (or in other words, a list of final states, saying for example that after all the decaying is done, you'll get 4 particles of this type arriving at these sorts of angles at this time, and then 3 particles of this other type arriving... etc). Each of these productions will have a different probability, and each decay channel/final state will have a different probability of resulting from its initial particle production.
So you try to match up the things your detector found, with these final states. Because so much of your model is based on probabilities, you have to do this statistically-- you have to measure a huge number of events, and then you measure whether the number of events of each type that you saw was close to the number of events of each type that your model predicts will occur on average. You ask, was the final tally of events closer on average to what the model tells us we'd see if no Higgs are being produced? Or is it closer to what the model tells us we'd see if the Higgs
was being produced? Or is something else entirely happening?
Does that all make sense? I am pretty sure if you look around you could find a more specific description of the decay channels that the LHC itself will be looking for. (Dorigo links
this site which supposedly contains some kind of catalog of decay channels for different kinds of particles, but I can't quite seem to find that data...)
