Publishing an AnalysisPublishing an Analysis
Physics collaborations such as those at large particle accelerator facilities at CERN or Fermilab are organic entities in themselves! Hundreds of minds gather together with the aim of producing results that the world has not seen before. Different cultures, different backgrounds and even different approaches help achieve the common goals in everyone’s sights.
It is important to understand how such a large group of people present their physics results to the rest of the community: A scientific paper written by a single person is the product of one mind; a paper by a collaboration the size of CMS is the product of over 2000 physicists.
Each paper goes through an exhaustive journey from early analysis to publication. This journey has several fixed waypoints, which exist so that every CMS member (who will sign the paper eventually) has the possibility to look at it. Everybody in the collaboration has the right and duty to contribute to the paper and must duly be informed when a paper is in production.
Beginning an analysis
As data for analysis starts to be collected from the LHC’s collisions, the life of a CMS paper typically begins with one of the many physics working groups such as those dedicated to studies of the Top quark, the Higgs boson, Supersymmetry, and so on. Each of these working groups has conveners who co-ordinate the analyses made by its members.
Physicists wishing to perform a particular analysis approach the conveners of the respective working group with a proposal in the form of a presentation. The conveners then decide whether to proceed with the analysis or not. If they decide to go ahead, the analysis topic gets added to the CMS Analysis Database and can be accessed via its Interface (CADI).
Any member of the collaboration can access CADI to find a list of the topics currently being analysed along with a list of authors responsible for them. The topics under analysis are not a mystery; it is known who the participants are, it is well known where more effort is needed and people are welcome to join.
The authors regularly present their studies at working group meetings and as the analysis makes progress it proceeds to the stage known as the pre-approval. The analysis must be ready for all members of the collaboration to see at least one week before the pre-approval meeting. The last leg before this pre-approval involves the Analysis Review Committee (or ARC), which is typically made up of three people appointed by the Physics Co-ordinator and the Publications Committee Chairperson.
Appointing a committee is not easy. The large collaboration has people with different skills and it is important that the right people be selected for the committees. Senior members are often asked to nominate others for this role.
The conveners decide to hold the pre-approval meeting when they are satisfied with the progress the authors have made. At this meeting, one of the authors presents their findings to other members of the collaboration.
Discussions during or after the presentation can sometimes get very heated. Yes, everyone might belong to the same team, but no punches are pulled when it comes to passing a critical eye over the work of one’s colleagues; perhaps this is so because everyone expects the papers to which they are signatories to be watertight. It is healthy that different people have different points of view.
While the working group is responsible for oversight of the analysis before pre-approval, after this crucial point the ARC assumes responsibility. When the authors have answered all the questions raised at the pre-approval and during the ARC review, and have made any necessary modifications, the analysis must be approved by the collaboration. The approval talk is announced two weeks before it is to occur. Preferably any discussion should happen before the meeting, not at the last moment. So, the documentation for the analysis is frozen and made available at the same time as the announcement so collaboration members can read it.
Documentation is important, including that which is only circulated internally. Ideally, other physicists should be able to reproduce the analysis themselves after reading the paper. Not everyone in the collaboration will look at all of it, but the level of scrutiny is sufficient that no paper produced would be rejected by a journal.
At the approval talk, CMS members ask questions after the presentation. The Physics Co-ordinator listens to questions and replies, and tries to understand if there are important points that need more scrutiny. The ARC then gives its opinion — to approve the analysis or do more work. After approval of the physics, if the paper as a whole is not ready but CMS would like to present analysis to a conference, a Physics Analysis Summary (PAS) is made available for distribution outside the collaboration.
There is another important stage before the paper is ready to be published. The draft of the paper is circulated for a collaboration-wide review (CWR) — which could happen in parallel with the analysis approval process or after it. In the CWR, signatories to the paper comment on text, and on the physics described in the text.
There is a distinction between these two, sometimes parallel, processes: While the physics approval is for CMS as a whole, the CWR includes all people who sign the paper — a set that overlaps with CMS members but includes others as well. There are a few people that sign the paper but who are no longer members of the CMS collaboration. When you join CMS, you start signing the papers one year after you have joined and you stop signing one year after you have left the collaboration. There are rules for people who have collaborated with CMS in the past, because work on the experiment started over 15 years ago.
After an analysis has been approved and the paper passes the CWR, the ARC, authors and others go over the entire paper in what is known as the Final Reading. This is the final stage of the paper before it leaves the collaboration and is done to ensure that all CMS papers communicate physics with clarity, have the same style and retain the same form. At the Final Reading, the paper is “blessed” by the Publication Committee, after which it is frozen and sent to a journal, where the CMS paper goes through an anonymous peer-review process, as would papers from smaller collaborations or individual scientists. Sometimes, there are suggestions from the referees although most papers are published without major changes.
- This text is a shorter version of an article that appreared in the CMS Times on 21st February, 2011, under the name Life of a CMS Paper.