OPT-IN: Measurement of Beam-Recoil Observables Cx and Cz for K+Λ Photoproduction

[David Ireland]

Dear Shankar et al, 

This is a very nice analysis and paper. 

Reinhard has already made a number of points that I would have mentioned, but won't repeat. 

Just a minor niggle that I appreciate is more of an analysis review comment. Recall that Cx and Cy are asymmetries of physical observables (=real numbers). Even though you are not measuring them directly, we know a priori that they must be bounded in [-1,+1]. Any data that appear outside these bounds (and there are a few for Cz at lower energies) are mathematically impossible, unless there is an experimental systematic error. Therefore, we can attribute the amount a data point is outside the physical limit as a measure of systematic error. In the small number of points that I noticed, this can be up to 0.5, which is an order of magnitude different to the estimates of systematics that are quoted in the paper. 

The kinematic fitting that is employed in the analysis uses the same logic: energy and momentum conservation force the kinematic quantities into a physical region of phase-space.

There are too few of the >1 points to make much of a difference to the theoretical model fits and intepretations, so I am not suggesting redoing anything. BUT,... there is a nightmare scenario in which ALL points are subject to the same systematic error that the points ouside the physical region are subject to and that the data points should ALL be lowered by 0.5! Unfortunately that overall difference would be at least as big as the differences between models, and would indicate that the data do not have any discriminating power. Unlikely, but not impossible!

Maybe a short comment could be added that the data points outside the physical region are indeed unphysical and are the result of statistical fluctuations in the extraction process.

Best wishes, 

Dave