| 1 |
|
\section{Limit on new physics} |
| 2 |
|
\label{sec:limit} |
| 3 |
|
|
| 4 |
< |
{\bf \color{red} The numbers in this Section need to be double checked.} |
| 4 |
> |
%{\bf \color{red} The numbers in this Section need to be double checked.} |
| 5 |
|
|
| 6 |
|
As discussed in Section~\ref{sec:results}, we see one event |
| 7 |
|
in the signal region, defined as SumJetPt$>$300 GeV and |
| 8 |
|
\met/$\sqrt{\rm SumJetPt}>8.5$ GeV$^{\frac{1}{2}}$. |
| 9 |
|
|
| 10 |
< |
The background prediction from the SM Monte Carlo is |
| 11 |
< |
1.4 $\pm$ 0.5 events, where the uncertainty comes from |
| 12 |
< |
the jet energy scale (30\%, see Section~\ref{sec:systematics}), |
| 13 |
< |
the luminosity (10\%), and the lepton/trigger |
| 14 |
< |
efficiency (10\%)\footnote{Other uncertainties associated with |
| 15 |
< |
the modeling of $t\bar{t}$ in MadGraph have not been evaluated.}. |
| 10 |
> |
The background prediction from the SM Monte Carlo is 1.3 events. |
| 11 |
> |
%, where the uncertainty comes from |
| 12 |
> |
%the jet energy scale (30\%, see Section~\ref{sec:systematics}), |
| 13 |
> |
%the luminosity (10\%), and the lepton/trigger |
| 14 |
> |
%efficiency (10\%)\footnote{Other uncertainties associated with |
| 15 |
> |
%the modeling of $t\bar{t}$ in MadGraph have not been evaluated. |
| 16 |
> |
%The uncertainty on $pp \to \sigma(t\bar{t})$ is also not included.}. |
| 17 |
|
The data driven background predictions from the ABCD method |
| 18 |
< |
and the $P_T(\ell\ell)$ method are 1.5 $\pm$ 0.9 and |
| 19 |
< |
$1.8^{+2.5}_{-1.8}$ events respectively. |
| 18 |
> |
and the $P_T(\ell\ell)$ method are $1.3 \pm 0.8({\rm stat}) \pm 0.3({\rm syst})$ |
| 19 |
> |
and $2.1 \pm 2.1({\rm stat}) \pm 0.6({\rm syst})$, respectively. |
| 20 |
|
|
| 21 |
|
These three predictions are in good agreement with each other |
| 22 |
|
and with the observation of one event in the signal region. |
| 23 |
|
We calculate a Bayesian 95\% CL upper limit\cite{ref:bayes.f} |
| 24 |
|
on the number of non SM events in the signal region to be 4.1. |
| 25 |
< |
This was calculated using a background prediction of $N_{BG}=1.4 \pm 1.0$ |
| 26 |
< |
events. The upper limit is not very sensitive to the choice of |
| 25 |
> |
We have also calculated this limit using a profile likelihood method |
| 26 |
> |
as implemented in the cl95cms software, and we also find 4.1. |
| 27 |
> |
These limits were calculated using a background prediction of $N_{BG} = 1.4 \pm 0.8$ |
| 28 |
> |
events, the error-weighted average of the ABCD and $P_T(\ell\ell)$ background |
| 29 |
> |
predictions. The upper limit is not very sensitive to the choice of |
| 30 |
|
$N_{BG}$ and its uncertainty. |
| 31 |
|
|
| 32 |
|
To get a feeling for the sensitivity of this search to some |
| 33 |
|
popular SUSY models, we remind the reader of the number of expected |
| 34 |
< |
LM0 and LM1 events from Table~\ref{tab:sigcont}: $6.5 \pm 1.3$ |
| 35 |
< |
events and $2.6 \pm 0.4$ respectively, where the uncertainties |
| 34 |
> |
LM0 and LM1 events from Table~\ref{tab:sigcont}: $8.6 \pm 1.6$ |
| 35 |
> |
events and $3.6 \pm 0.5$ events respectively, where the uncertainties |
| 36 |
|
are from energy scale (Section~\ref{sec:systematics}), luminosity, |
| 37 |
< |
and lepton efficiency. |
| 37 |
> |
and lepton efficiency. Note that these expected SUSY yields |
| 38 |
> |
are computed using LO cross-sections, and are therefore underestimated. |
| 39 |
|
|
| 40 |
< |
In Figures XX and YY we provide the response functions for the |
| 41 |
< |
SumJetPt and \met/$\sqrt{\rm SumJetPt}$ cuts used in our analysis, |
| 42 |
< |
{\em i.e.} the efficiencies of the experimental cuts as a function of |
| 43 |
< |
the true quantities. Using this information as well as the kinematical |
| 44 |
< |
cuts described in Section~\ref{sec:eventSel} and the lepton efficiencies |
| 45 |
< |
of Figures~\ref{fig:effttbar}, one should be able to confront |
| 46 |
< |
any existing or future model via a relatively simple generator |
| 47 |
< |
level study by comparing the expected number of events in 35 pb$^{-1}$ |
| 48 |
< |
with our upper limit of 4.1 events. |
| 40 |
> |
Conveying additional useful information about the results of |
| 41 |
> |
a generic ``signature-based'' search such as the one described |
| 42 |
> |
in this note is a difficult issue. The next paragraph represent |
| 43 |
> |
our attempt at doing so. |
| 44 |
> |
|
| 45 |
> |
Other models of new physics in the dilepton final state |
| 46 |
> |
can be confronted in an approximate way by simple |
| 47 |
> |
generator-level studies that |
| 48 |
> |
compare the expected number of events in 35 pb$^{-1}$ |
| 49 |
> |
with our upper limit of 4.1 events. The key ingredients |
| 50 |
> |
of such studies are the kinematical cuts described |
| 51 |
> |
in this note, the lepton efficiencies, and the detector |
| 52 |
> |
responses for SumJetPt and \met/$\sqrt{\rm SumJetPt}$. |
| 53 |
> |
{LOOKING AT THE 38X MC PLOTS BY EYE, THE FOLLOWING QUANTITIES LOOK ABOUT RIGHT.} |
| 54 |
> |
The muon identification efficiency is $\approx 95\%$; |
| 55 |
> |
the electron identification efficiency varies from $\approx$ 63\% at |
| 56 |
> |
$P_T = 10$ GeV to 91\% for $P_T > 30$ GeV. The isolation |
| 57 |
> |
efficiency in top events varies from $\approx 83\%$ (muons) |
| 58 |
> |
and $\approx 89\%$ (electrons) at $P_T=10$ GeV to |
| 59 |
> |
$\approx 95\%$ for $P_T>60$ GeV. |
| 60 |
> |
{\bf \color{red} The following quantities were calculated |
| 61 |
> |
with Spring10 samples. } |
| 62 |
> |
The average detector |
| 63 |
> |
responses for SumJetPt and $\met/\sqrt{\rm SumJetPt}$ are |
| 64 |
> |
$1.00 \pm 0.05$ and $0.94 \pm 0.05$ respectively, where |
| 65 |
> |
the uncertainties are from the jet energy scale uncertainty. |
| 66 |
> |
The experimental resolutions on these quantities are 10\% and |
| 67 |
> |
14\% respectively. |
| 68 |
> |
|
| 69 |
> |
To justify the statements in the previous paragraph |
| 70 |
> |
about the detector responses, we plot |
| 71 |
> |
in Figure~\ref{fig:response} the average response for |
| 72 |
> |
SumJetPt and \met/$\sqrt{\rm SumJetPt}$ in MC, as well as the |
| 73 |
> |
efficiency for the cuts on these quantities used in defining the |
| 74 |
> |
signal region. |
| 75 |
> |
% (SumJetPt $>$ 300 GeV and \met/$\sqrt{\rm SumJetPt} > 8.5$ |
| 76 |
> |
% Gev$^{\frac{1}{2}}$). |
| 77 |
> |
{\bf \color{red} The following numbers were derived from Spring10 samples.} |
| 78 |
> |
We find that the average SumJetPt response |
| 79 |
> |
in the Monte Carlo is very close to one, with an RMS of order 10\% while |
| 80 |
> |
the response of \met/$\sqrt{\rm SumJetPt}$ is approximately 0.94 with an |
| 81 |
> |
RMS of 14\%. |
| 82 |
> |
|
| 83 |
> |
%Using this information as well as the kinematical |
| 84 |
> |
%cuts described in Section~\ref{sec:eventSel} and the lepton efficiencies |
| 85 |
> |
%of Figures~\ref{fig:effttbar}, one should be able to confront |
| 86 |
> |
%any existing or future model via a relatively simple generator |
| 87 |
> |
%level study by comparing the expected number of events in 35 pb$^{-1}$ |
| 88 |
> |
%with our upper limit of 4.1 events. |
| 89 |
> |
|
| 90 |
> |
\begin{figure}[tbh] |
| 91 |
> |
\begin{center} |
| 92 |
> |
\includegraphics[width=\linewidth]{selectionEff.png} |
| 93 |
> |
\caption{\label{fig:response} Left plots: the efficiencies |
| 94 |
> |
as a function of the true quantities for the SumJetPt (top) and |
| 95 |
> |
tcMET/$\sqrt{\rm SumJetPt}$ (bottom) requirements for the signal |
| 96 |
> |
region as a function of their true values. The value of the |
| 97 |
> |
cuts is indicated by the vertical line. |
| 98 |
> |
Right plots: The average response and its RMS for the SumJetPt |
| 99 |
> |
(top) and tcMET/$\sqrt{\rm SumJetPt}$ (bottom) measurements. |
| 100 |
> |
The response is defined as the ratio of the reconstructed quantity |
| 101 |
> |
to the true quantity in MC. These plots are done using the LM0 |
| 102 |
> |
Monte Carlo, but they are not expected to depend strongly on |
| 103 |
> |
the underlying physics. |
| 104 |
> |
{\bf \color{red} These plots were made with Spring10 samples. } } |
| 105 |
> |
\end{center} |
| 106 |
> |
\end{figure} |
| 107 |
> |
|
| 108 |
> |
|
| 109 |
> |
|
| 110 |
> |
%%% Nominal |
| 111 |
> |
% ----------------------------------------- |
| 112 |
> |
% observed events 1 |
| 113 |
> |
% relative error on acceptance 0.000 |
| 114 |
> |
% expected background 1.400 |
| 115 |
> |
% absolute error on background 0.770 |
| 116 |
> |
% desired confidence level 0.95 |
| 117 |
> |
% integration upper limit 30.00 |
| 118 |
> |
% integration step size 0.0100 |
| 119 |
> |
% ----------------------------------------- |
| 120 |
> |
% Are the above correct? y |
| 121 |
> |
% 1 16.685 0.29375E-06 |
| 122 |
> |
% |
| 123 |
> |
% limit: less than 4.112 signal events |
| 124 |
> |
|
| 125 |
> |
|
| 126 |
> |
|
| 127 |
> |
%%% Add 20% acceptance uncertainty based on LM0 |
| 128 |
> |
% ----------------------------------------- |
| 129 |
> |
% observed events 1 |
| 130 |
> |
% relative error on acceptance 0.200 |
| 131 |
> |
% expected background 1.400 |
| 132 |
> |
% absolute error on background 0.770 |
| 133 |
> |
% desired confidence level 0.95 |
| 134 |
> |
% integration upper limit 30.00 |
| 135 |
> |
% integration step size 0.0100 |
| 136 |
> |
% ----------------------------------------- |
| 137 |
> |
% Are the above correct? y |
| 138 |
> |
% 1 29.995 0.50457E-06 |
| 139 |
> |
% |
| 140 |
> |
% limit: less than 4.689 signal events |
