1: %auto-ignore
2: %\section{Data samples}
3:
4:
5: The present study uses data from $p\bar{p}$ collisions at $\sqrt{s} = 1.8$
6: TeV collected by the Collider Detector at Fermilab (CDF)\cite{CDF} during the
7: period 1992-1995 (Run I). The integrated luminosity of the data sample is 109
8: $\pm$ 7 pb$^{-1}$. Events were selected \cite{toplj,topdi} and assigned to
9: three different $t\bar{t}$ subsamples chosen for their low background and
10: high efficiency for $b$ jet identification. Each sample is classified by the
11: number of leptons and identified $b$ jets in the final state.
12: %The $t\bar{t}$ data samples are obtained from an integrated luminosity of 109
13: %$\pm$ 7 pb$^{-1}$ of $p\bar{p}$ collisions at $\sqrt{s} = 1.8$ TeV collected
14: %from Run I (1992-1995) at the Tevatron and observed
15: %in the CDF detector \cite{CDF}. Three sub-samples of $t\bar{t}$ data
16: %previously used in the measurement of the top quark mass~\cite{toplj,topdi}
17: %are chosen for their low background and high efficiency for $b$ jet
18: %identification.
19:
20: The ``dilepton'' sample is dominated by $t\bar{t}$ in which both $W$ bosons
21: decay to an electron or muon and neutrinos. Events are selected by
22: requiring $\MET >$ 25 GeV, one muon and one electron of opposite charge with
23: $P_T > 20$ GeV in the central pseudo-rapidity region ($|\eta| < 1.0$) \cite{eta}, and two jets with
24: $E_T > 10$ GeV and $|\eta| < 2.0$. This is a subsample of the
25: dilepton events used in other analyses~\cite{topdi}, considering only $e +
26: \mu + jets$ events in order to remove the dominant background,
27: which is Drell-Yan production of $ee$ or $\mu\mu$.
28: %Using only $e\mu$ dilepton events, which is expected to
29: %account for 58 $\pm$ 2 \% of the total dilepton event acceptance, has
30: %an advantage since in the standard top dilepton analyses, reducing
31: %background from 2.4 $\pm$ 0.5 to 0.76 $\pm$ 0.21, while reducing the
32: %expected signal from 4.7 to 2.8.
33: The significant remaining backgrounds are decays to electron and muon of $Z
34: \to \tau \tau$, $WW$ in association with extra jets, and $W$ production
35: associated with three or more jets, where one jet is misidentified as an
36: electron or a muon. No attempt is made to identify $b$ jets explicitly.
37: However, initial and final state gluon radiation can result in extra jets, so
38: the $b$ jets are assumed to be the two highest $E_T$ jets, which is correct
39: in $\sim$80 \% of dilepton events. There are four $M_{\ell b}$ combinations in
40: each dilepton event.
41:
42: The other two samples used in the analysis require only one $W$ to decay into
43: an electron or muon and a neutrino and the other $W$ to decay hadronically
44: (``lepton+jets''). These events are selected by requiring one electron or
45: muon with $P_T >$ 20 GeV, in the central region as above. At least four jets
46: are required, three of which must have $E_T >$ 15 GeV, $|\eta| < 2.0$, and
47: the fourth must have $E_T >$ 8 GeV and $|\eta| < 2.4$. The background for
48: these events consists predominantly of direct production of a $W$ plus extra
49: jets and its behavior is modeled with the VECBOS generator \cite{VECBOS}. To
50: reduce the background, at least one jet must be identified as a $b$ candidate
51: ($b$-tagged) with a topological algorithm requiring tracks in the jet
52: reconstructed with the silicon vertex (SVX) detector to form a secondary
53: vertex \cite{toplj,oldtopresults}. This requirement is 48 \%
54: efficient for tagging at least one $b$ jet in a $t\bar{t}$ event
55: \cite{cdf_mass_prd}. Without any $b$-tag, the expected signal to background
56: ratio ($S/B$) of the sample is 0.4, whereas requiring one $b$-tag improves
57: $S/B$ to 5.3. The $b$-tag also selects the jet to be paired with the lepton
58: to form $M_{\ell b}$. Events with a single $b$-tagged jet comprise the
59: ``single-tagged'' sample, and have one measured $M_{\ell b}$ which is correct
60: half the time. Events with both $b$ quarks tagged make up the
61: ``double-tagged'' sample, have a $S/B$ of 24, and provide two $M_{\ell b}$
62: pairings, at least one of which combines the wrong $b$ with the $\ell$.
63:
64: A total of 7 events were found in the dilepton $e \mu$ sample
65: with an expected background of 0.76 $\pm$ 0.21 events. In the
66: single-tagged sample 15 events were found with a background 2.0 $\pm$
67: 0.7, and in the double-tagged sample there were 5 events with a 0.2
68: $\pm$ 0.2 background. Note that since right-handed
69: leptons have higher $P_T$, an increase in events passing the lepton
70: $P_T$ trigger requirement could also indicate a V$+$A theory. However,
71: any potential observed rate increase would be deemed to be {\em a posteriori}
72: knowledge from the point of view of this analysis, and therefore only the
73: shape of the $M^2_{\ell b}$ distributions is considered.
74:
75: