Statistics data analysis is one of those things that experimental physicists learn along the way. It is not a topic usually included in the curriculum studiorum of physics students at Universities: only few basic ingredients are taught during laboratory courses, and not much is added to that during a typical Ph.D. program.
One usually learns the most common tools to fit histograms, combine measurements, estimate uncertainties on the field, as these things are always needed to produce publishable physics results. But several key statistical concepts often remain fuzzy and obscure in the mind of a large fraction of experimental physicists throughout their career. I know this because this happened to me, too - for quite a few years after my graduation.
One of the hardest hurdles in one's education is realizing how little does one know about a subject. Once that realization is achieved, the stimulus toward learning more is automatic. I started studying statistics and as I made progress I found out I had been a fool not to do that earlier in my career. After a few years of studies, I can at least say I know very well how little do I know about different subjects...
With that realization came a wish to distribute more knowledge about statistics among my peer. This is presently one of my main occupations within the CMS collaboration, in fact. And when last fall my colleague Luca Lista contacted me -along with Giovanni Punzi and Riccardo Faccini- to put together a scientific committee with the goal of organizing a one-week school about Statistics for high-energy physicists, I was of course delighted.
(Picture above: the participants busy during an exercise session)
Even more delighted I am now, as I just returned from that week of courses that were held in Vietri sul Mare, a very beautiful seaside town on the Amalfi coast, just south of Naples. The school was a real success: we brought there 100 students who followed with attention a total of 15 hours of frontal lessons plus 6 hours of exercise sessions.
If you want to give a look at the slides shown at the school please visit the conference web site here (the slides are in the timetable page).
The speakers were highly knowledgeable and very good lecturers: Glen Cowan (who wrote a very useful book on the topic of the school), Luc Demortier, Jochen Ott, Harrison Prosper, and Lorenzo Moneta. They covered a broad range of topics, focusing on the most important issues for particle physics data analysis, including confidence limits, multivariate analysis, and the usage of the most advanced packages we use for producing results (roostats and roofit).
We also got a half-day off on Wednesday, with two possibilities for an excursion: a tour of the coast, or a visit to the archaeological site of Pompei. I chose the latter, and spent four hours in the ancient town - a real jump back of 2000 years. Pompei was buried on AD79 by the pyroclastic ashes from the eruption of the nearby volcano Vesuvio, and it has remained almost perfectly preserved to the modern era, when it was re-discovered.
(Above, me and my fiancee in front of the ruins of the foro)
Will we organize another such school next year ? I hope so, although it might be a better idea to wait two years instead. On the other hand, the applications exceeded our availability this time, so we left out about 20 students. It should not be hard to get 100 more applicants next year...
The INFN School Of Statistics In Vietri, 2013