Here’s a summary of results for null-test calculations of single-spin asymmetries for the Run 5 charged pion A_{LL} analysis. Asymmetries are defined as (Y,B):

Yellow Beam: -1 x { (++ + +-) - (-+ + —) }
Blue Beam: -1 x { (++ + -+) - (+- + —) }
Like-Sign: (++) - (—)
Unlike-Sign: (+-) - (-+)

All the asymmetries incorporate final beam polarizations.

We do see some deviation from a null result (~2 sigma) in the pi+ Yellow Beam and Unlike-Sign SSAs. The combination suggests an excess of counts in the (Y,B)=(-,+) state. However, the lack of any corresponding effect in the pi- sample leads me to believe this is just a statistical fluctuation.

Update 2008-03-11

Removed runs that used bxing pattern 111. Murad found that the relative luminosities for this pattern are screwy. His work is documented here, here, and here. The second link has the most details about bxing pattern 111.