[Qweak_transverse_prl_comments] Fwd: [Q-weak] First draft of the Qweak elastic transverse asymmetry paper

[Buddhini Waidyawansa]

---------- Forwarded message ----------
From: Greg Smith <smithg at jlab.org>
Date: Tue, Sep 16, 2014 at 5:10 PM
Subject: Re: [Q-weak] First draft of the Qweak elastic transverse asymmetry
paper
To: Buddhini Waidyawansa <buddhini at jlab.org>


 Hi Buddhini,

I have some comments... First though, congratulations! I was really
surprised how well written your paper was right out of the gate. Excellent
job! Thank you so much for doing this, for all of us. I know better than
most how much work it is to write a draft. Thank you for all the hard work
you put into this! I really enjoyed reading it.

After Adesh's thesis, I expected to find the word "the" either missing or
used when it should not be on every line of this paper, like it was
(almost) on every line of his thesis. Whew!

I am a really picky critic. It's amplified by my unfamiliarity with the
subject of your paper, but the flip side of that is that that makes me a
great critic, one that's more representative of your target readership. So
some of my questions may be dumb. But if I have them then so will most
readers I reckon. SO please don't get upset by my comments, which I put
forward in the spirit of trying to help make the paper better. Don't take
them personally- that's 1000% not what they should be misconstrued as. I
just want to help, and I think you did an outstanding job already. You said
you only had a few comments so far, and so I read your manuscript carefully
in the hopes that it will make the 2nd draft much closer to the final
draft! I'm saying all this because when I got comments on the Wien 0 PRL
like what I'm sending you now, I wasn't happy for the most part. I wanted
my drafts to be perfect (they weren't!). I hope that's not your reaction
here.

I'll pass you my marked-up paper copy now- it has things not included
below. It's also attached.

In general, and even though I made the same mistake sending out the draft
of our commissioning PRL for comments, I think it would be better to have a
larger font (for us old guys), double spaced,  line-numbered version (as
well as a companion version in the final PRL format to check length, etc.)
for editing drafts. It's almost impossible to comment on a paper version of
the draft you sent, it's too closely spaced. And line numbers make your job
easier as well as the reviewers, something I learned during the NIM editing
process!

First: hyphens. I am hyphen-blind and was mercilessly slammed by Katherine
Myers over this often-ignored aspect of English grammar. She must have
pointed out at least 100 problems of this type in the NIM paper. Google
"hypenation rules" and you'll see what the rules are, but also you'll see
here
<http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0CCoQFjAB&url=http%3A%2F%2Fwww.apastyle.org%2Flearn%2Ffaqs%2Fwhen-use-hyphen.aspx&ei=hzMXVLTvNteqyASK24L4Aw&usg=AFQjCNELOuFjf_5DQsf-TaDBoXjyXTl8Gg&bvm=bv.75097201,d.aWw>
that the rule is pretty slippery and it's hard to know when to use it and
when not to. In a nutshell, if there are 2 adjectives before a noun, you
hyphenate them unless it's common usage not to, but especially when doing
so helps improve the meaning even in the slightest. In the abstract for
example, I think "Beam-Normal" should be hyphenated. "Electron-Proton" and
"electron-nucleon" too in front of the noun "Scattering". "helicity-flip"
in front of "amplitudes". But I don't know if "Single-Spin" should be
hyphenated in front of the noun "Asymmetry" or not (probably... ask
Katherine). These appear multiple times throughout the paper. And there
will be many more missing hyphens in other examples throughout the paper. I
won't elaborate further.

OK- just spoke to you about this, and you suggested focusing mostly on the
big picture for this 1st draft. I am marking up a paper copy, with
difficulty, but here are some big picture comments:

I don't think we have ever discussed this as a collaboration, or on the
council, but I wonder whether or not on some of the ancillary papers where
1 or 2 individuals made outsized contributions, they (you) should be 1st
author. I like that you went alphabetical for now. I think we should
definitely keep that for the main Qweak papers, where so many individuals
contributed. But for this one, as well as others that may come down the
pipe, I wouldn't object to putting your name 1st (on this one), and then
going alphabetical after your name. We'll have to discuss it- it's a bit of
a hot potato topic I guess. I'll try to remember to bring it up (better I
do, or someone else does,  than you).

In the middle of the 1st column, page 2, you say both that there is "recent
disagreement" and "good agreement" between the Rosenbluth and pol. xfer
methods. You can't have it both ways.

At the end of the 1st column, page 2, it gets very confusing. You refer to
the well-known 7.7 sigma p-rad problem. Then an 11% correction. Then a 1%
effect on the correction. I can't tell what the percentages are percentages
of. You have to lay out this landscape in a single language, not 3
different ones. How much of the 7.7 sigma problem is the proton
polarizability term, and how much of that is up for grabs due to the
uncertainty in the TPE corrections? If it's 1% of an 11% correction that in
turn is only a small part of the 7.7 sigma discrepancy, then I don't see
how this is relevant, in practice. If it's a significant piece, then the
situation is completely different and much more relevant as a motivation
for the results presented in your paper. But from the numbers you provide I
can't tell. If your primary motivation for our BNSSA measurement is a
completely irrelevant correction to the p-rad puzzle, it's
counter-productive. If it's in fact relevant, and can potentially amount to
a non-trivial part of the p-rad discrepancy, then that has to be made
crystal clear.

Has someone checked the formulas? (I haven't!).

In the 1st sentence of the 2nd paragraph in the 2nd column on p. 2, you
define BNSSA as when the pol. is normal to the scattering plane. How can
you do this when we have an azimuthally symmetric detector with essentially
8 different scattering planes?

Run 1 vs Run 2: This comes up a few times but I really think it's confusing
jargon and should be avoided in a short paper like this. Some other way of
distinguishing the 3 groupings of measurements should be done, or maybe
nothing should be done. The fact they are separated in time is not really
relevant and could just be mentioned once, casually, and then you just deal
with 1 horizontal and 2 vertical measurements.

"Data set": You use this for slugs, for runs, for Run 1 and Run 2, for H &
V, in other words, you use it for everything and never define it. It's
jargon. It needs to go. Sorry.

Bottom of page 3: You make the 1st reference to "3 data sets" which put me
off balance. I only realized later that you are presenting the 2 V and 1 H
measurements- I think this needs to be made clear up front, before you
refer to "... the three data sets."

Top of page 4: I would recommend you spend the effort to make it 100% clear
what you are presenting as your main result in Table 1:  You say it's the
avg of the fits- what does that mean? I have no idea. Both fits? Each fit?
3 fits? Even if I assume you mean each fit, the asymmetries vary from -4 to
+4 ppm in Fig. 1. You fit them with Eq. 3. What is presented in Table 1,
labeled "physics asymmetries"? It can't be A_det from Eq. 3 because A_det
depends on phi_det. I don't know (in principle) where the numbers in Table
1 come from or what they could possibly be. You absolutely have to clearly
define what you mean by A_exp.

The fits: they should be explained much more clearly. What was floated in
the fit, and what was not? Was P fixed? To what? Was B_n varied (I assume
it was, and that it's the mystery result you present as A_exp in Table 1).
But what about what you call the "floating phase offset" phi_det? That's
not floated, it's known for each detector. Is there some other thing called
phi_det you actually do float in the fit? If so what is it, what value does
it have, is it 1 floated variable in your fit or 8? Also, what is the
"floating constant"? Is that also 8 variables or one? What value comes out?
You only have 8 points to fit- it's important to spell out exactly how many
free parameters are floated in the fit and what they are. It seems to me
there should be only 2: B_n and phi_s. And phi_s is the fudge factor you
talk about as both the "floating phase offset" and "floating constant".

Finally, only 2 curves show up in Fig. 1. What about the second set of
results with V polarization?

The 2nd sentence  on p. 4 seems out of place and confusing: an "acceptance"
of 49% of 45 degrees? Is that technically an acceptance? The word "only" in
front of 49% bugs me too. And what does this have to do with anything
anyway? I don't get it. I certainly don't see the connection (if that's
what you are trying to do here) to the next sentence: The 3rd sentence on
p.4: Are you referring here to the correction for going from <A(Q^2)> to
A(<Q^2>)? That was R_Det=0.980 in Wien 0. What the heck is this 0.9938
correction? Finally, 4 digits seems extreme, whatever this correction
actually is. Is it? Maybe not.

middle of 1st column, p4: Could you elaborate a little more how the
inelastic bkg f and A were measured?

When putting a negative sign in front of a number like 5.0, always use
$-$5.0 because the math mode minus is bigger. Don't leave a space between
the - sign and the number, either.

Your factors that go into R_tot are all different from our Wien 0 numbers.
Is this a mistake? I'm not aware that this was revisited.

You should explain how you go from the 3 mystery numbers in Table 1 to the
single B_n you report 1/3 of the way down the 2nd column on p. 4. I assume
you did a weighted average somewhere.

Relative error table: relative to what? To B_n?

I was surprised not to find any discussion about how well the calculations
do with other data out there- doesn't G0 have a result? Perhaps you don't
have the results of the predictions at the G0 kinematics. But I seem to
recall there was something.

I assume you don't have error estimates for the other 2 calculations?

Finally, there is a lot of space in the intro building up the connection to
the p-radius and magnetic moment business. But at the end I don't see a
firm connection. Only a soft one. Can you make a more quantitative
statement about the impact this has on the 7.7 sigma puzzle for example? Or
on the ingredients to that puzzle? It seems to fall a bit short of
expectations otherwise. It certainly stands on its own, however, with
respect to what we learn about the importance of these multi-pion
excitations. That's very solid. If the p-radius connection is really so
soft, perhaps it would be better to de-emphasize that to some extent in the
intro and build up the connection to testing aspects of the calculations
and learning more about the TPE  process instead.

References: I did not check them yet. Somebody should eventually.

OK, again, thanks a zillion for all the work you put into this. It's a
great job!

Greg



-- 
-------------------------------------------------------------------
Buddhini Waidyawansa
Postdoctoral Fellow
C122,
12000 Jefferson Ave,
Newport News, VA 23602.
TP 757-912-0410
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.jlab.org/pipermail/qweak_transverse_prl_comments/attachments/20141111/c552c3ef/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: img-916165827-0001.pdf
Type: application/pdf
Size: 3349904 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/qweak_transverse_prl_comments/attachments/20141111/c552c3ef/attachment-0001.pdf