I once was an active chessplayer, but work duties have long taken tournaments off my plate - I simply do not have the time to sit through long hours of chess battles. So I play blitz online on chess.com (my handle is "tommasodorigo", in case you wondered).
Last week I spent a few interesting days in the pleasant winter setting of Engelberg, a mountain location in the Swiss alps. There I attended the CHIPP 2012 winter school, an event organized by Vincenzo Chiochia from University of Zurich and Gabriella Pazstor from University of Geneva. They invited me to give a three-hour mini-course in Statistics for data analysis in High-Energy Physics, something which was a new experience for me. It took me the best part of the last couple of months to get prepared, but I was glad I did. In the end, the material I put together could have been used profitfully for five or six hours of lecture, but by skipping some of the topics I could get to the end without using a silly speed.
The CMS Collaboration has just released the results of a deep study of their sample of lead-lead collisions, produced at a center-of-mass energy of 2.76 TeV per nucleon by the Large Hadron Collider.
This week's edition of the CMS Times features a short piece by A.Rao, where some points are made on the issue of correct statistical analysis of high-energy physics data.
...and people who like sausages, should not ask how they are made.As a member of two large scientific collaborations (CDF and CMS), I enjoy the benefit of seeing lots of scientific publications that carry my name as an author being produced at weekly rates. This is however also a burden, since I at the very least must try to ensure that I like the way the results are produced. I.e., that I agree with the details of how these scientific measurements are made.
Update: I have modified the title of this post [originally: "Opera's Statistical Booboo"] and the text below (in places I marked accordingly) upon realizing, thanks to a very good point raised by a reader in the comments thread below, that the idealization I was making of the measurement described below made my conclusions too hasty. Read the text to the end if you want more detail.
A blog is by nature a place where things move on fast. Articles disappear beyond the horizon in the matter of a week or two, and only rarely get resuscitated by a later article linking them back from oblivion. At Science 2.0 things are no better than in any other blog sites, with the aggravating feature that there is no "archive" button, nor a "random post" feature. Since I believe that many of my articles are not very connected to the specific time at which they have been written, I have in mind to reorganize the material somehow, when I have the time. However, this looks like a grievious task, since the number of posts I have written here is about 500.
Just as you thought it was over for 2011, and you proceeded to hung the Higgs mass plots on the christmas tree as a wish for stronger signals next year, ATLAS comes out with a new particle discovery. That's what I like of particle physics - there's always so much going on that the excitement is never really over.
NASA is featuring the first high-resolution images of asteroid Vesta as taken by the DAWN spacecraft during a low-altitude orbit. The images show a very interesting surface, battered with old and more recent craters, plus "textures such as small grooves and lineaments that are reminiscent of the structures seen in low-resolution data from the higher-altitude orbits. Also, this fine scale highlights small outcrops of bright and dark material." (from the NASA piece).
Have you ever looked at a histogram with the data displayed as counts per bin in the form of points with error bars, and wondered whether those fluctuations and departures from the underlying hypothesized model (usually overimposed as a continuous line or histogram) were really significant or worth ignoring ?
Have you ever looked at a histogram with the data displayed as counts per bin in the form of points with error bars, and wondered whether those fluctuations and departures from the underlying hypothesized model (usually overimposed as a continuous line or histogram) were really significant or worth ignoring ?
Particle physics experiments usually invest a considerable part of the time used to produce a measurement in the task of determining the corresponding uncertainty on the estimate, or -when a new effect is observed (say a quantity is measured away from zero, when zero would be the "null hypothesis", predicted by the current model)- estimating the statistical significance of the observation.
Another confirmation of correct evaluation of controversial HEP measurements awaited me just after the Higgs evidence was presented at CERN. I am sort of embarassed by this compulsory self-promotion, but this is my blog so I will excuse myself ;-)So the story is the following... Some of you might still remember the controversy over the Omega b discovery, at the Tevatron a couple of years ago.
