[Clascomment] OPT-IN: Electromagnetic decay of the ??*0 to ???

[Reinhard Schumacher]

Hello Dustin,
	With a fresher head, I reread the part of the paper that I said was 
confusing.  This time I see what you were trying to say, and it seems 
pretty reasonable.  For clarity I would recommend changing one sentence:

page 9, column 2, paragraph 6, last 3 lines:  "To maximize COUNTING 
statistics WHILE minimizing THIS uncertainty, the optimal..."  This gets 
rid of the offending word 'recovery' in quotes, which I thought may be 
needlessly distracting.

My other comments still apply, of course.

Best Regards,
Reinhard


___________________________________________________________________
Reinhard Schumacher         Department of Physics, 5000 Forbes Ave.
Carnegie Mellon University, Pittsburgh, PA 15213, U.S.A.
phone: 412-268-5177         web: www-meg.phys.cmu.edu/~schumach
___________________________________________________________________


Dustin Keller wrote:
> Hi Reinhard,
> 
> Thank you for your careful consideration and suggestions on our paper. 
> The first suggestions you mentioned are mostly easy fixes.  In your last 
> comment you mentioned that the discussion of the uncertainty when 
> compensating for events lost in doing the two confidence level cuts was 
> unclear.  The correction for events lost is inherent in the ratio 
> calculation seen in Eq. 21 for the simplified two channel case.  In the 
> original study of the confidence level cuts optimization Eq. 21 was used 
> to calculate the resulting ratio from a Monte Carlo mixture of  
> radiative and pi0 decays.  The recovery uncertainty was minimized while 
> considering both the level of statistics seen in the data and the 
> statistical uncertainty produced in the ratio.  After using these 
> studies to find a range of optimized confidence level cuts for P^a and 
> P^b a variation of these cut was chosen that would lie just outside the 
> optimization range for each cut.  The cut and the effect of the 
> variation is listed in Table 3.  For presentation clarity the numerical 
> specifics are omitted and only a general statement is given however the 
> CLAS-note 2010-015 goes into this type of study and is available for the 
> general public.  We give the appendix to help guild a motivated reader 
> through the calculation of the full ratio.  So with the acceptance terms 
> in Table 2 and the resulting raw counts from the fits in each case one 
> quickly knows the corrections for
> each channel.  What maybe confusing in the discussion you mentioned is 
> the phrase "recovery" uncertainty.   This is the difference between the 
> ratio input to the Monte Carlo and the value recovered using the 
> described analysis framework with specific confidence level cuts.  We 
> would be interested in clearing up any confusion with the hope of 
> conservatively adding to this discussion as to not dilute the more 
> critical points.
> 
> thanks again,
> Dustin
>