Do you remember the dijet mass bump found by CDF in W plus jets events ? That signal, whose significance exceeded four standard deviations, had everybody around go crazy for a while. Theoretical models appeared overnight in the arxiv, experimentalists canceled trips to spin their data to confirm the signal, others speculated on wild systematic uncertainties creating the effect, bloggers placed bets on the reality of the effect.

Time has passed, and the bubble of course has deflated a little. The signal has not been conclusively been disproven yet, but there are indications that it might not be genuine after all. First of all, DZERO made a similar search, showing that they did not see anything anomalous in the mass region where CDF was observing a nice bump. Now, ATLAS has also produced a study of the dijet mass distribution of pairs of jets produced together with a well-identified leptonically-decaying W boson. It is the one shown below.

The data appears to be very well modeled by backgrounds, and in fact the lower part of the figure showing the ratio between data and background prediction has points which wiggle around 1.0 without departing significantly from there, and mostly staying within estimated systematic uncertainties (the light brown shading). Does this mean that the CDF signal is again on "life support" (to steal a nice catchphrase used by David Gross at EPS, where he discussed the CMSSM as being threatened to be fully excluded by present searches) ?

Well, not so fast. As you can see, the ATLAS data contain a smaller contribution from WW/WZ events as the CDF selection; this is a result of the higher centre-of-mass energy of the LHC collisions creating more backgrounds to the electroweak diboson production processes than at the Tevatron. The WW/WZ signal (which correspond to a W or a Z boson decaying into a pair of jets, and thus being reconstructed in the surroundings of 80-90 GeV) appears in the ATLAS plot as a tiny contribution (in white), and the experimental data points (in black) actually appear to ignore it, following all too happily the sum of all other backgrounds, except for one single point which stands in more agreement with the backgrounds plus diboson prediction.

So what ? Well, if the WW/WZ signal is unobservable in the present ATLAS search, why should we conclude that the W+X CDF signal should be visible ? In fact, if the latter were present and if it had the same relative rate with respect to WW/WZ as it does at the Tevatron, it would not be observable at all: the region between 130 and 150 GeV has too much background for a signal smaller in rate than the WW/WZ one. So the figure does not deliver, after all. Still, it is quite nice to see these studies progressing our understanding of QCD production of jets associated with vector bosons!

For more information on the ATLAS analysis see here.