Banner
    Apology to my ATLAS Colleagues
    By Tommaso Dorigo | July 10th 2012 05:30 AM | 13 comments | Print | E-mail | Track Comments
    About Tommaso

    I am an experimental particle physicist working with the CMS experiment at CERN. In my spare time I play chess, abuse the piano, and aim my dobson...

    View Tommaso's Profile
    I realize I owe an apology to my ATLAS colleagues.

    In a post-Higgs-announcement article where I was listing some random "post-mortem" considerations, I wrote a paragraph which could be read as an accusation of malpractice in one of the recent ATLAS analyses. I was misled by the fact that I had no access to the details of the
    actual results being shown at ICHEP on the matter -my information was partial and I was not sure whether it was public or not; it was second-hand and I had no way to check its real source.

    When I wrote that paragraph I was aware that its contents could be potentially damaging to ATLAS, but on the other hand I thought it was important information that was worth pointing out. So I wrongly outweighted the need of being a gentleman with my colleagues across the ring with the journalistic loyalty to my readers here. That was a lapse of good judgement. Furthermore, it transpires that it is quite possible that the information I distributed is not even correct.

    In the same paragraph of that article I did write very clearly that I fully acknowledged the genuine result of ATLAS, but that was probably not enough. So I apologize to my colleagues who felt harmed by my posting. I have now removed the offending part from the article.

    Being a blogger and being part of a scientific collaboration are two almost fully incompatible things. It is sometimes excessively hard, when not downright impossible, to be both loyal to one's colleagues -or even one's competitors, as in this case- and write honestly about one's research. But I am also convinced of the importance of distributing information on basic research through this means of communication. I have not drawn logical conclusions from these facts yet.

    Comments

    Holding off for a day or two, to verify information, is not incompatible with blogging. :)

    Could you convince someone from ATLAS for a guest post here, so we can hear the other side, too?
    They can maybe give us (and yourself) some details you did not have access to.

    i don't see this as a problem. you got the rumors out there, and that prompted a response, which cleared up the situation, and now its sorted out, and you issued an apology. ideally everyone acts like grownups in this situation and moves on.

    hopefully a couple of people from ATLAS give you their phone numbers so that you can call them up ahead of time in the future.

    i don't really see this as any kind of huge tension between the role of a blogger and the role of scientific collaborator, think you're confusing things and missing the deeper fact that both are entirely human, and therefore flawed endeavors that are not amenable to precision theories.

    dorigo
    Ah Lamont, I wish things were half as naive as you picture them !
    As if calling a colleague from the other experiment would clear things up ?! We cannot
    release any internal information, including confirming or denying anything.
    Sorry if I cannot explain this one well to you here now, but as far as understanding the
    subtleties in this issue you're like on Mars or something.

    Cheers,
    T.
    The Stand-Up Physicist
    Will we ever know when you knew the Higgs at 125.3 GeV had at least 4 sigma of support?  Is revealing that kind of info problematic?
    dorigo
    Hi Doug,

    4 sigma were there since last December if you eyeballed a combination of all LHC information, as shown by Philip Gibbs a long time ago. That is what I called "firm evidence" back then.

    Other than that, yes, that is internal information. People are just crazy with their attempts at keeping information reserved, so I cannot even tell you whether a top group meeting was held at 4PM or 5PM yesterday. Forget about 4sigma or 5sigma.

    Cheers,
    T.
    Oh, nice move Tommaso! I see you just deleted my comment from yesterday night. Are you trying to hide evidences? :-) What a shame, I thought you better than that...

    dorigo
    John, your comment included the information that ATLAS considered offensive. I tried to edit it out but for some reason could not, and am in the middle of a meeting.... You should have been more careful.

    Please repeat your question (I know there was one, but forgot now).

    Cheers,
    T.
    I was just asking why you felt like reporting a rumor that you acknowledge as potentially false, and was clearly offensive to the scientific integrity of your competitors. I mean, this was a clear case where one checks her sources very carefully before publishing, otherwise one immediately gets sued (and I guess this is more or less what happened, given your apology post. Am I right?). So, what was your goal there? Honestly, the suspect you were deliberately trying to discredit your competitors was very strong, since I cannot see how your audience could have profited from that false information, with that inexistent backing.

    dorigo
    I won't go in the details, but I was indeed not sure whether what I was reporting was official or not (and thus the thing had therefore some explanation, thus not necessarily offensive to the competitors). This due to the fact that other than the CERN seminars there was no other information I could access, since I was unable to reach the ICHEP material. The information, if true, would have been useful to put in context the speculations on the discrepancies of LHC findings with SM Higgs predictions. That was in fact the purpose of discussing in the same context the early top cross section measurements by my former experiment, CDF. But I do agree that I should have avoided discussing the issue completely.

    Cheers,
    T.

    Good of you to apologise.

    Hi T,

    Atlas shows a significance of around 4.6 sigma in its digamma channel while the expected is 2.4 sigma. This is as per Figure 29 of ATLAS-CONF-2012-091 . So isn't there an exclusion of the standard model Higgs at 95% confidence ( 2 sigma) due to the excessive excess (2.2 sigma) in a small sliver of region near 265.5 GeV. Shouldn't this region be mentioned as a STANDARD MODEL Higgs exclusion at 95% CL in the Atlas report? In the abstract and conclusion the 95% exclusions are mentioned but this small but significant region is not included in that!!

    Am I missing something or is Atlas not mentioning what it should?

    Also I don't understand why its diphoton signal is mentioned as 4.5 sigma by Atlas in its abstract while both the graph in Fig 29 and Vixra Higgs tool put it at 4.6 sigma.

    Atlas Report is at: http://cdsweb.cern.ch/record/1460410/files/ATLAS-CONF-2012-091.pdf

    Ravi

    The significance as read off the exclusion plot is only an approximation and I am happy that I get as close as 4.6 if the true value is 4.5. Accurate figures require the p-value plots which take into account the non-Gaussian nature of the liklihood disributions.