Ab VTX-Q and Ac Vtx+K summary PRL Jan/04 vs summer-01 ------------------------------------------------------ Ab Vtx-Q: PRL Summer01 Measurement = 0.919 0.9208 stat error 0.018 0.0184 err % error val (Ab=0.919) AP calib 1.41 0.0130 0.0133 b purity 0.63 0.0058 0.0068 Note 1 Rc 0.11 0.0010 0.0010 Rb 0.29 0.0027 0.0027 Ac 0.05 0.0005 0.0005 Ab 0.00 0.0000 - trk eff 0.34 0.0031 0.0021 Note 3 trk res 0.04 0.0004 0.0016 Note 3 pi misid 0.00 0.0000 0 QCD Cf 0.35 0.0032 0.0032 QCD Sf 0.31 0.0028 0.0029 c-AP b-tag 0.56 0.0051 0.0052 g->cc 0.01 0.0001 0.0001 g->bb 0.02 0.0002 0.0003 eff uds 0.01 0.0001 0.0001 Araw uds 0.09 0.0008 0.0009 same hemi cc 0.01 0.0001 0.0001 same hemi bb 0.21 0.0019 0.0020 cc Ec correl 0.14 0.0013 0.0012 bb Eb correl 0.10 0.0009 0.0006 pol Pe 0.50 0.0046 0.0048 MC stat 0.34 0.0031 0.0031 total 1.89 0.0174 Ac VTX+K: PRL Summer01 Measurement = 0.673 0.6734 stat error 0.029 0.0290 err % error val (Ac=0.667) AP calib 2.96 0.0199 0.0210 c purity 0.68 0.0046 0.0030 Note 4 Rc 0.28 0.0019 0.0019 Rb 0.30 0.0020 0.0020 Ac 0.00 0.0000 - Ab 0.10 0.0007 0.0007 Note 2 trk eff 0.36 0.0024 0.0010 Note 2 trk res 0.49 0.0033 0.0024 pi misid 0.12 0.0008 0.0021 QCD Cf 0.18 0.0012 0.0012 QCD Sf 0.59 0.0040 0.0040 c-AP b-tag 0.83 0.0056 0.0057 g->cc 0.22 0.0015 0.0017 g->bb 0.06 0.0004 0.0005 eff uds 0.13 0.0009 0.0008 Araw uds 0.43 0.0029 0.0031 same hemi cc 0.33 0.0022 0.0025 same hemi bb 0.04 0.0003 0.0003 cc Ec correl 0.48 0.0032 0.0032 bb Eb correl 0.07 0.0005 0.0006 pol Pe 0.50 0.0034 0.0035 MC stat 0.64 0.0043 0.0043 total 3.48 0.0234 Notes on changes: 1) In the final Ab analysis, the b-purity error is calib. stat only. The IPxy tail of 3% at 100micron included in summer01 was excessive and the magnitude for a more realistic tail is negligible. Some other very small b production/decay systematics looked like just MC stat fluctuation are also dropped from the b-purity error. 2) Small dAb,dAc due to c,b fragmentation , variations used at summer01 (dAb:+-0.0001,+-0.0008; dAc:+-0.0007,+-0.0003) are dropped. The relevant momentum correlation estimate already covers fragmentation effect. The effect of varying the mean on correlation is mostly MC stat. fluctuation. 3) Tracking efficiency and resolution were reevaluated since summer01. The resolution smearing/z-shift used latest prescription, while the tracking efficiency correction used the vertex charge distribution width itself is as the constraint. The Qvtx contraint is a better procedure as the standard tracking inefficiency was derived from just the total multiplicity which depends on many generator level multiplicity modeling discrepancies as well as real tracking reconstruction ineffiecieny simulation. Because the Qvtx is a sum of charges, it only depends on heavy fragmentation model, lifetime and heavy hadron decay multiplicity (mostly, except the charm hadron 1-prong fraction for ccbar events), at second order, while removed dependency on light flavor fragmention where the model reliability is more doubtful. However, this procedure added some sensitivity to the charged vs neutral heavy hadron production fraction. This procedure also absorbed the VXD-alone vector reweighting (data/MC discrepancy) systematic in summer01. The track resolution systematic followed revised common recipe recommended for all analyses. 4) Summer 01 result lumped c-purity calib stat error into combined AP calibration to avoid confusion for correlating the tag purity with other analysis. This is now split off and c-purity is purity calib stat. error, while the IPsmear and c prod/decay errors previously in summer01 are dropped. At summer-01, an error for XY IPsmear of 3% @ 100 micron tail was included to give dAb=0.0024 and dAc=0.0013. In the last IP check YT study by Sean for Rb,Rc, the 97-98 data showed no sign of long XY IP tail (< ~0.2% level) while 96 data have a tail roughly 0.5%@100 micron. Given the small weight of 96, the overall effect of IP tails looks indeed negligible.