From: Glen Warren (gwarren_at_anti-spam)
Date: 14-May-2003 11:19:22
GEn01 analysis meeting May 7, 2003
====================================
attended by: Paul, Oscar at UVa
Donal, Hongguo, Frank, Nikolai and Glen at JLab
REPORTS
*) GMN: Glen used the Basel Gmn paper which has a fit to get Gmn. Donal
directed him to Jim Kelly's paper on the neutron form factors. For our
Q2, there is little difference between the two. We will either take the
average of the two and a suitable error, or just use dipole. There is
some justification for using dipole.
Another issue is how Gmn affects the asymmetry in Arenhoevel's model.
According to Hongguo's thesis, a larger Gmn increases AVed, which is
opposite of what one would expect for AVed ~ Gen/Gmn. Hongguo is not
able to determine if their was a mixed in the files. Donal will look at
that. If he cannot resolve this puzzle, we will ask Arenhoevel for a
new point with Gmn proportional to but not equal to Dipole.
*) ypos: we discussed exactly what the offset should be. Hongguo was
worried that the wrong sign to the ndet angle offset was used, but we
believe it to be correct. The argument is that MC shows that the peak
of the distribution is close to 53.3 degrees. The data shows a peak at
negative ypos. This means that the ndet must be moved to a larger angle
in the engine so that the peak of the distribution is centered around
53.3, as in the MC.
This means that the theta_np^cm and ypos alignment approaches provide a
consistent picture!
*) W vs. theta: Glen reviewed his findings in which he examined the
theta dependence of W. Early he had seen such a dependence. Now, that
dependence is largely gone. The difference is the use of symmetric
Gaussian (earlier) and asymmetric Gaussian (now) fits. There is not
enough of a dependence to worry about a theta dependence W cut.
*) W cut: Nikolai asked why are we doing a W cut at all. His point was
that if there is a problem with W, then we should be very careful about
apply cuts. He claimed that the agreement between data and MC is very
good, so dilution factor will not be compromised by removing the cut.
The two standard reasons that are given for the W cut is cut
optimization and reduction of region to a kinematic region in which we
expect the MC and data to agree. Some of the latter is done for us by
the E' cut, so there is some redundancy there.
It the end, we concluded we would leave the cuts as is for now. If
someone has a better approach later, then we will pursue it.
*) protons: There was a lot of discussion about how to use the proton
asymmetries. Nikolai advocates using the proton asymmetries as a basis
for setting kinematics. He argues that the statistics are better and
the asymmetry larger. Glen is opposed to using the protons for anything
other than diagnostic purposes. He argues that ultimately we are not a
proton experiment and worries that there might be unaddressed issues
that might impact the protons. It was also pointed out that the
nitrogen asymmetry contribution will depend on the dilution factor, so
any comparison of protons asymmetries to MC should include this
correction kinematic dependent correction. Donal also pointed out that
the proton asymmetry will also depend on how Arenhoevel parameterized Gep.
For Q2=0.5, the measured proton asymmetry is within 3% (3 sigma) of the
"expected" value, which is within our systematics.
*) 98 data and different engines: There is good agreement for the 98 NH3
data for the 98 and 01 engine. We believe that they are using the same
HMS central momentum, and that it is unlikely that the offsets were
used. One question raised was if the field setting code had changed
between 98 and 01.
*) Comparison of 2nd and 3rd pass: Glen will use the fine binned butcher
results to make a better comparison.
*) background: Frank looked over how many events are gained and lost.
The problem is that his time stamp technique cannot tell him if events
gained were also lost. As the background is on the 0.5-1% level, this
is most important in understanding how good events might be lost be the
extra background, which would impact the asymmetry differently than as
just a straight dilution factor. Glen suggested that assume two cases:
new events have neutron asymmetry and new events have no asymmetry.
Take difference as uncertainty.
TO DO
I'm not bothering with a to do list this time as the minutes are going
out a couple of hours before the next meeting. I thought I had already
submitted these notes, but in my sleep-deprived stupor of last week I
forgot to.
This archive was generated by hypermail 2.1.7 : 14-May-2003 11:20:13 EDT