ODSL Forum: "CLs – What to report when you know your hypothesis test will have limited sensitivity"


Presenter: Alexander Read, University of Oslo
Title: CLs – What to report when you know your hypothesis test will have limited sensitivity

Abstract: A large number of searches for new phenomena in particle physics are reported each year, most of which are a null result, i.e., with no evidence for a new phenomenon in the data. Unlike in many other fields, we feel that these null results can be as important scientifically as headline-grabbing discoveries for increasing our knowledge. CLs was invented at the time of the search for the Higgs boson at LEP (1989-2000), as the collision energy increased from the mass of the Z0 boson to over the threshold for producing Z0-pairs, and the question of whether we had seen the Higgs boson or not in each new dataset became particularly urgent. One of the goals was to have a way of quantifying results of searches for new signals, e.g., the Higgs boson, that both had optimal sensitivity to a real signal and yet did not take “unfair” advantage of background fluctuations to set stringent confidence limits on new phenomena that an experiment in fact had little or no sensitivity to.


Further reading:

  1. Modified frequentist analysis of search results (The CL(s) method), a contribution to Workshop on Confidence Limits in 2000 at CERN:
  2. Presentation of search results: The CL(s) technique, a contribution to PHYSTAT 2002 in Durham:

As usual, the format will be a short presentation followed by plenty of discussion.

Please let all interested know about the Journal Club (get them to send their Email address).  We are looking forward to lively discussions.