It was suggested I turn my attention to the third p_T bin (6 < p_T < 7 GeV/c) where we observe a large systematic in a region with better statistics. Here, I post follow-up plots similar to used to investigate the high-x_F bin discussed in the Spin-PWG meeting. I believe the conclusion is that this is also a statistical fluctuation.

Figure 1: Current Money Plots

For reference sake I post the current status of the asymmetry measurement.

Figure 2: High-mass Band Asymmetries (p_T for x_F < 0)

Raw	Background Corrected

Both the raw and background-corrected high-mass band asymmetries are consistent with zero. At 6 < p_T < 7 GeV/c the raw mass window and raw high-mass band differ by 1.00σ. The background-corrected mass window and high-mass band differ by 1.04σ (though, now events are correlated to the extend the same polarization and signal fit is used, integrated over non-overlapping mass ranges).

Figure 3: Systematics Summary (p_T for x_F < 0)

One can see from the systematics summary plot that in the third bin the high-mass band asymmetry is ~1σ away from the mass-window asymmetry. This suggests that the size of the difference is due to statistical fluctuation. Note: the high-mass band contains the majority of the data excluded by the tight window. It does not contain all of these data, however, as there is one bin at the low end still excluded. Such a thing should not change the message, but it should be kept in mind.

Figure 4: Asymmetries for 6 < p_T < 7 GeV/c (x_F < 0)

Raw	Background-corrected
0.1 < mγγ < 0.2 GeV/c2
0.105 < mγγ < 0.165 GeV/c2
0.165 < mγγ < 0.2 GeV/c2

In Fig. 4 I post the asymmetries as a function of φ_S for both raw asymmetries (left) and background-corrected asymmetries (right). One can compare the asymmetry changes bin-by-bin. I note that the fit quality for the high-mass band is not very good (~2.67σ). This is improved to 2.10σ when one moves to the single-arm measurement (Fig. 5). The high-mass band and tight-window φ_S cross-ratios differ by 0.50σ, 1.56σ, 0.38σ, 2.02σ, 1.86σ, and 0.92σ, in order of increasing φ_S-bin.

Figure 5: Single-arm Asymmetries (High-mass Band)

Figure 6: Mass Distributions for 6 < p_T < 7 GeV/c (x_F < 0)

In Fig. 6 I post the spin-sorted mass distributions for the third p_T bin. One can see by eye the distribution of events which fall into the various mass windows.