Wed Oct 11 12:41:31 CDT 1995
from run 1A and 1B.
The analysis is identical to the
published dijet search paper using
run 1A data alone, but here we also check QCD predictions for the mass and
angular distribution.
The dijet mass spectrum
is compared to a smooth fit, a QCD and CDF detector simulation normalized to
the data in the first six mass bins, and an excited quark and CDF detector
simulation. Like the
inclusive jet Et spectrum, the
dijet mass spectrum on a linear scale
exhibits a smoothly increasing excess over QCD which becomes
most noticeable at high mass: the measured
spectrum is more energetic than QCD predicts. However, the
angular_distributions
agree with QCD alone, and constrain the region with mass greater than 625 GeV
to have less then 13% isotropic angular distribution at 95% CL (in this
region the excess over QCD on the
mass plot is 47%). It is
therefore unlikely that the excess above QCD at high dijet mass and high
jet Et is due to new physics with an isotropic angular distribution.
Any physics explanation for the excess would need to have a dijet angular
distribution similar to QCD. One possible explanation, a larger than
expected gluon distribution of the proton, is being investigated by the CTEQ
collaboration.
The search for new particles uses the data itself to determine the background
by simply fitting the data to a smooth parameterization.
The comparison between the mass data and the background fit on a
logarithmic scale and
linear scale in bins equal
to the mass resolution shows no
significant evidence for new particles, which would appear as an upward
fluctuation in 2 or mass bins. Fitting the data to both a smooth
parameterization of the background and a simulation of the mass
resonance, we find upper limits on the cross section for new particle production
vs. mass.
The 95% CL upper limits on signal
cross section are compared to lowest order models of new particle production
using CTEQ2L parton distributions. Where the model's cross section is higher
than the measured upper limit, we have an
excluded mass region, which we
compare to exclusions by other experiments.
This preliminary search excludes at 95% CL models of: