91av

EVEN physicists get their sums wrong. A glimpse of the eagerly sought
particle known as the Higgs boson, reported last year, was revealed last week to
be just a mathematical slip.

Scientists at CERN in Geneva reported seeing signs of the elusive Higgs last
September, just as their LEP accelerator was due to be demolished to make way
for a more powerful machine. In a flurry of excitement, they persuaded their
managers to let them run LEP for an extra month to gather more data
(91av, 9 September 2000, p 4).
The Higgs boson is believed to endow
other particles with mass, and physicists are desperate to find it.

Researchers studying data from collisions in two of LEP’s four detectors
found a bigger signal than they expected in the area where the Higgs should
appear. This signal stuck out above the noise by as much as three standard
deviations—meaning there was 0.3 per cent chance it was due to random
noise. CERN policy says you need five standard deviations to claim a discovery,
though three is strong evidence. With an extra month’s data, the evidence was
not much stronger and LEP was closed, to the anger of many physicists.

Now a re-analysis of last autumn’s results shows that some of the
calculations were wrong. “At the beginning of November they were doing analysis
very quickly,” a CERN spokesman told 91av.

The new analysis shows researchers overestimated the collision energy of the
particles, so the level of noise relative to signal was higher than expected.
The correction means the relative size of the excess signal is just two standard
deviations, giving it a five per cent chance of being due to noise. That’s still
enough to be tantalising, says Chris Tully of Princeton University, who is
responsible for data analysis on one of the detectors. But it’s not enough for
LEP to go down in history as the accelerator that discovered the Higgs, he says.