Blind Analysis

I'm a member of the MiniBooNE collaboration, and we are doing a blind analysis in our search for neutrino oscillations. The idea is that we wish to prevent biasing ourselves before we complete the analysis of the data. We are doing a "closed box" blind analysis, which means that we sequester the events that appear to be signal-like and do not perform any analyses on them before the analysis chain is complete.

Our analysis is a search for electron neutrino events in a muon neutrino beam, which is the signature of neutrino oscillations seen by the LSND experiment. We are performing the experiment to confirm or rule out the LSND neutrino oscillation result. Effectively, the blind analysis means that we use other data samples, like the muon neutrino data and cosmic muon decay electron events, to understand our event reconstruction and analysis algorithms. We do not use electron neutrino events that might come from neutrino oscillations in the development of the analysis, but only after the algorithms are complete. We are currently in the final stages of the analysis, and are hoping to open the box soon, although we have been saying that for a while!

We chose to do a blind analysis for many reasons, but one of the key reasons is that the LSND result, if it is due to oscillations, would be inconsistent with the Standard Model's prediction of only three families of neutrinos. Thus, the LSND result has huge ramifications if it is confirmed and we deemed it necessary to use the most strict methods in our search for these oscillations.

As a MiniBooNE collaborator, I am often asked if we will see a signal, which is to say: do I think the LSND signal is real? I've come to realize I don't care if we see a signal. All that matters to me is getting it right. Frankly, I think it would be scientifically irresponsible for me to hold a strong opinion about it one way or the other. I think that the beauty of science comes from the idea that Nature can reveal her secrets if we ask the right questions and are open to the answers. Approaching this analysis with a strong bias one way or the other would be tantamount to closing one's mind to a certain type of answer, and to me that would be a failure.

A lot of people in the field feel that the signal is false, and that we will rule out LSND-type oscillations, with almost religious conviction. It belies their bias in what should be an objective pursuit. I think part of it stems from the saga of the 17 keV neutrino. In that case the scientific method was vindicated (although I am sure that there was plenty of subjective chatter amongst the participants, especially at conferences) but I think the experience left a lot of people in the field uncomfortable with new and different experimental results in neutrino physics. We will shortly learn whether or not the LSND result was a false alarm, or whether Nature is a lot more complex than we thought.

And I can't wait to know the answer, whatever it might be!