Things that go bump in the night

DZero tried to faithfully reproduce a CDF result that had the blogosphere buzzing in early April. As is evident from the picture, no bump was observed corresponding to the one seen by CDF. The corresponding CDF figure can be seen here.

Things that go bump in the night can send a chill up your spine. They are unexpected, unknown and could be anything. Bumps in data have a very similar effect on physicists. If we see an unexpected bump in a data plot, it could be the very first observation of something unknown. A discovery might be in the offing.

On April 6, CDF announced observation of an unexpected bump in its data. An update of this result was shown just two weeks ago. In events in which two jets were made at the same time as a W boson, CDF researchers wondered what the mass of the parent would be if the two jets were the decay products of a single, parent particle. They found that there was an unexpected excess at a mass of about 145 times heavier than a proton. When CDF collaborators announced their result in a seminar at Fermilab, the news spread like wildfire across the blogosphere and even made the front page of CNN.

With such a flurry of excitement, it was imperative for the observation to be confirmed or disproved. The only other experiment that could duplicate this exact measurement is the DZero experiment and, naturally, we immediately looked at our data. While our similar result already existed and was immediately available for scrutiny, the two experiments had made different choices on data selection. In order to reproduce the CDF analysis, DZero needed to modify its analysis to duplicate choices made by the CDF collaboration. Further, with a possibly-momentous discovery on the line, it was critical that DZero do the analysis extremely carefully. Confirmation or rejection of the bump that sent shivers up the spines of millions of science enthusiasts is a serious business.

Last Friday, the DZero experiment released its measurement, which reproduced the CDF methodology as closely as possible. As you can see in the figure above, the collaboration did not observe a similar bump.

So where does that leave us? Experts will continue to debate how DZero could have missed a real effect or how CDF could have inadvertently manufactured a signal. A task force created by Fermilab Director Pier Oddone and the CDF and DZero collaborations will coordinate a study of the two experiments’ analyses. Over the next weeks and months, the picture will be clarified and the truth revealed. This is the beauty of the scientific method.

Don Lincoln

These physicists performed this very delicate analysis. It is the official position of the DZero collaboration that the analyzers for this analysis were selected solely on the basis of talent, and their names, while appropriate, are purely coincidental.
Fermilab’s Mike Diesburg has the very important task of processing all of the DZero data to make it suitable for physics analysis. He is one of the very few people who makes contact with every event recorded by DZero.