A spreadsheet with calculations of the chisq / d.f. and associated probability for the A_N results of J. Drachenberg is attached.  The first set of 6 points corresponds to x_F < 0 and 6 < p_T < 7 GeV/c, and is in response to questions raised at the recent spin PWG meeting.  The second set of one point is for the largest negative x_F bin (integrated over all p_T).

For each of the 6 + 1 points, there are 4 pairs of counts.  The four are different spin states (up, down) and hemispheres of the detector (up, down).  The pairs correspond to wide (0.1 < IM < 0.2 GeV) or narrower (0.105 < IM < 0.165 GeV) invariant mass windows.  Ratios of counts for the two windows are examined for fluctuations significantly outside expectations.