RevModPhys.84.1477】Tests of the standard electroweak model at the energy frontier

By applying data derived mistag probabilities to

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ency to these events. The initial heavy-flavor fraction is derived from ALPGEN and PYTHIA simulation and is corrected by a factor of 1:4 Æ 0:4 derived from a jet control data sample. The b-tagging efficiencies are determined from simulation and checked using control data samples. The W þ jets background yields after b tagging in the D0 result are fixed by normalizing the simulated events to the untagged W þ jets data after subtracting the other backgrounds from the data. The relative contributions of the " W þ lf , W þ cc, and W þ bb in the untagged sample are fixed to the cross-section ratios predicted by MCFM. The yields in the b-tagged sample are then computed by applying flavor-based (mis)identification probabilities to the jets in simulated events. The probabilities (one for u-, d-, and s-quark initiated jets, one for c-quark initiated jets and for b-quark initiated jets) are derived using data control samples (Scanlon, 2006) and are parametrized as functions of jet pT and . The small background from multijet events is difficult to simulate accurately, so this component is also determined from control data samples. For CDF, a control sample is selected using events which have nonisolated leptons and low ET , and the yield in the signal sample is determined by 6 extrapolating the yield from this sample into the signal region having isolated leptons and high ET . For D0, this background 6 is determined by selecting a multijet dominated control sample with kinematics similar to the WH events, and then applying a probability that these events would be misidentified and appear in the signal sample. The multijet control sample is selected by requiring lepton candidates which pass very loose isolation requirements, and the background yield is computed by applying the event-by-event probability that these loose-isolation events would pass the standard isolation requirement and thus appear in the signal sample. Table XXXI shows the data yields and background and signal predictions for the CDF analysis, and Table XXXII shows the yields for the D0 analysis. The predicted Higgs yield includes not only WH events, but also a small contribution from ZH ! ‘‘bb events in which one of the leptons The predicted and observed yields for the CDF 2 fbÀ1 WH search separated by the detector region. From Aaltonen et al. Pretag events b tagging Mistag Wbb Wcc " tt(6.7 pb) Single top (s-ch) Single top (t-ch) WW WZ ZZ Z !  non-W QCD Total background WH signal (120 GeV=c2 ) Observed events ST þ ST Central region 32242 ST þ JP ST þ NN ST þ ST 3:88 Æ 0:35 37:93 Æ 16:92 2:88 Æ 1:25 19:05 Æ 2:92 6:90 Æ 1:00 1:60 Æ 0:23 0:17 Æ 0:02 2:41 Æ 0:26 0:06 Æ 0:01 0:25 Æ 0:04 5:50 Æ 1:00 80:6 Æ 18:8 0:85 Æ 0:10 83 11:73 Æ 0:92 31:15 Æ 14:03 7:87 Æ 3:43 15:56 Æ 2:39 5:14 Æ 0:75 1:87 Æ 0:27 0:93 Æ 0:11 1:84 Æ 0:20 0:08 Æ 0:01 1:29 Æ 0:20 9:55 Æ 1:73 87:0 Æ 18:0 0:60 Æ 0:07 90 107:1 Æ 9:38 215:6 Æ 92:34 167:0 Æ 62:14 60:68 Æ...
View Full Document

This document was uploaded on 09/28/2013.

Ask a homework question - tutors are online