1 |
\clearpage
|
2 |
|
3 |
\section{Systematic Uncertainties in Signal Acceptance}
|
4 |
\label{sec:syst}
|
5 |
|
6 |
In this section we discuss systematic uncertainties in the signal acceptance. These efficiency
|
7 |
uncertainties are relevant for the interpretations in the \wzmet\ and the GMSB models, which
|
8 |
are combined with the results of the trilepton and quadlepton analysis, respectively, in AN-2012/351.
|
9 |
|
10 |
\begin{table}[htb]
|
11 |
\begin{center}
|
12 |
\footnotesize
|
13 |
\caption{\label{tab:syst} Summary of uncertainties in the signal efficiency. }
|
14 |
\begin{tabular}{l|c|l}
|
15 |
\hline
|
16 |
\hline
|
17 |
Source & Value (\%) & Method \\
|
18 |
\hline
|
19 |
Luminosity & 4.4 & official CMS value \\
|
20 |
Trigger efficiency & 3 & efficiency measurements documented in Sec. 4, Table 9 of AN-2012/248 \\
|
21 |
Lepton ID/isolation & 2 (per lepton) & Z tag-and-probe measurements in AN-2012/257 \\
|
22 |
B-veto & 6 & dedicated measurement in AN-2012/248 Sec. 7.6 \\
|
23 |
Z mass window requirement & 3 & see text and Table~\ref{tab:mllsyst} \\
|
24 |
Jet selection, dijet mass, \MET & assessed at each model point & official JetMet POG recipe \\
|
25 |
\hline
|
26 |
\hline
|
27 |
\end{tabular}
|
28 |
\end{center}
|
29 |
\end{table}
|
30 |
|
31 |
A summary of the efficiency uncertainties is presented in Table~\ref{tab:syst}.
|
32 |
The CMS uncertainty in the luminosity is 4.4\%. The trigger efficiency is measured in AN-2102/248 with an uncertainty of 3\%.
|
33 |
The lepton identification and isolation requirements are measured in data and MC and found to be consistent within 2\%, for
|
34 |
the \pt\ $>$ 20 GeV region relevant for this analysis, in AN-2012/257. The impact of the b-veto on the signal acceptance
|
35 |
is quantified with a dedicated measurement performed in Sec. 7.6 of AN-2012/248. The uncertainty in the selection of dilepton
|
36 |
events satisfying the Z mass window requirement 81--101 GeV is performed as follows. In both data and MC, the Z mass window
|
37 |
in the inclusive preselection region (Z and at least 2 jets) is loosened to 60--120 GeV.
|
38 |
The efficiency of the events in the loose window to satisfy the analysis dilepton mass selection
|
39 |
of 81--101 GeV is compared in data and MC, and found to be consistent within 3\% for both ee and $\mu\mu$ channels
|
40 |
(see Table~\ref{tab:mllsyst}), and a corresponding uncertainty on the signal efficiency is assessed.
|
41 |
|
42 |
The above uncertainties are the same for all SUSY model points. However the impact of the jet energy scale uncertainty, which
|
43 |
affects the selection efficiencies for all jets and \MET\ objects, varies significantly across the model parameter space and
|
44 |
is assessed separately at each point. The official JetMet POG recipe is used for this purpose. Each jet is assigned an uncertainty
|
45 |
based on its \pt\ and $\eta$. The jet energy is varied by this uncertainty, which is propagated to the efficiencies for the jet selection,
|
46 |
dijet mass selection and \MET\ selection. In addition, for the \MET, a 10\% uncertainty on the unclustered energy is included.
|
47 |
The \MET\ variation alters the shape of the signal \MET\ distribution and causes a bin-to-bin migration of events, which is
|
48 |
included in the limit setting procedure performed with LandS.
|
49 |
|
50 |
\begin{table}[htb]
|
51 |
\begin{center}
|
52 |
\footnotesize
|
53 |
\caption{\label{tab:mllsyst} Summary of the dilepton mass selection efficiency uncertainties. Loose and tight refer to dilepton
|
54 |
mass windows of 60--120 GeV and 81-101 GeV, respectively.}
|
55 |
\begin{tabular}{l|c|c}
|
56 |
\hline
|
57 |
\hline
|
58 |
& ee & $\mu\mu$ \\
|
59 |
\hline
|
60 |
MC loose & 190661.9 & 251210.1 \\
|
61 |
MC tight & 174162.5 & 229648.0 \\
|
62 |
MC tight/loose & 0.913 $\pm$ 0.005 & 0.914 $\pm$ 0.004 \\
|
63 |
\hline
|
64 |
data loose & 209540 & 263747 \\
|
65 |
data tight & 185555 & 234132 \\
|
66 |
data tight/loose & 0.886 $\pm$ 0.003 & 0.888 $\pm$ 0.003 \\
|
67 |
\hline
|
68 |
\hline
|
69 |
\end{tabular}
|
70 |
\end{center}
|
71 |
\end{table}
|