[Eg1_run] Fwd: Re: Next EG1 INclusive meeting

[kuhn]

 From Rob

-------- Original Message --------
Subject: 	Re: [Eg1_run] Next EG1 INclusive meeting
Date: 	Tue, 29 May 2012 21:02:52 -0400 (EDT)
From: 	Robert G. Fersch <fersch at server1.pa.uky.edu>
To: 	kuhn <kuhn at jlab.org>



Hi -

     I included a .txt file containing responses to unaddressed points in
the analysis note for the committee (which I think should be read by the
group).  They are all fairly minor points, and it doesn't look like they
are requesting any changes to the note itself, so I am content to let that
situation be unless instructed otherwise....
     A final treatise on systematic errors and a table of data for archive
submission should be available soon.
     I am very unlikely to be able to phone in since I am a shift leader in
the day shift in STAR, though I will attempt to escape for a few minutes
to phone in if things are calm enough (there is a trainee who can take the
helm for me in that case....)

-Rob



On Tue, 29 May 2012, kuhn wrote:

>  Dear EG1ers,
>
>  we have our next meeting as usual tomorrow, Wednesday 5/30. at 11:00 a.m. in F226. Primary item of business: The most
>  recent response from the PWG committee and our answer. Also, progress reports on all other EG1-related analyses are
>  welcome.
>
>  - Sebastian
>
>  Call-in instructions:
>
>   1. First dial the conference call facility:
>       1. In US, participant dials the toll free number: 866-740-1260
>       2. International participant dials: 303-248-0285
>   2. Then enter the passcode 8156718, followed by the # sign
>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.jlab.org/pipermail/eg1_run/attachments/20120530/0d7ea95b/attachment.html 
-------------- next part --------------
Responses to the boldfaced points:

I don't understand this reply: the TOF v_eff calibration should be done BEFORE cooking with RECSIS. I agree that it is probably a minor effect, but still I don't understand why this step, that is normally done for all TOF calibration, has been skipped, in such a thorough analysis as this one.

It is going back a way, so this response is my best recollection of the issue:  

I am not sure the data banks required to reconstruct the terms needed to calibrate the effective velocity were available at the time, and my own understanding of the issue (at the time I completed these calibrations) was not thorough enough to rectify it, or to understand its comparative importance. In any case, the calibration was not done, and short of restarting from where we were in 2004 with this analysis, simply cannot be done now.



Page 126, fig. 3.12: can you show additional plot of the elastic peak position as a function of momentum, theta and phi. The integrated plots are not enough to judge the quality of the corrections.

No reply here?

Sorry for the accidental omission. These plots are indeed included in the new Figures 3.8, 3.9 and 3.10.



Has this been done? Are errors really unaffected by this fix?

Yes, this error has indeed been fixed; the errors were unaffected by any noticeable amount, so I didn't address this any further (though I should have clarified that this issue is resolved, as you suggest.)



Page 281, section " Radiative correction errors": did you include the errors on the radiative corrections due to the target model?

No reply here but it is included in the new version of the note

Since any changes in the radiative correction due to the target model aren't leading order effects on the asymmetry, a variation of these effects was accounted for by changing the ammonia target length and regenerating the radiative corrections in a single iteration of the data, as detailed in the note.

Page 306, third paragraph: can you show a plot with the comparison of A1 and A2 extracetd from the data with the resenbluth separation and the model prediction, and the corresponding Rosenbluth separation plot for one of these critical cases? If the errors on the A_par points used for the Rosenbluth sepration and the corresponding error on the fit intercept and slope would be correct, than the extracted values of A1 and A2 may be far from that predicted by the model or (for A1) from what extracted using a model for A2 but they should still be compatible within the errors. Is this the case? If not it could indicate that the errors are wrong.

No reply?

If I am understanding this comment correctly, this is an issue that should be resolved with using the data to iteratively redefine the model, as we have done, and that the requested plot is represented by any A1, A2 plots that are shown with the model for comparison in the final results. I think this addresses the point you request...


Page 330, Fig. 8:30, what does weighting with Q^2 add to weighting with x ?. g_2 has a higher-twist part which should contribute with 1/Q.

No reply?

I agree now, this isn't a very useful plot, and it will not be included as a final result in the paper. (I originally included it as a parallel to existing plots for g1 used in the duality paper, that aren't really relevant to g2.)