# Top Quark Mass Measurement in the Dilepton Channel Using the Full CDF Data Set

## Authors:

• Julian Budagov1
• Igor Suslov1
• George Velev2
• 1JINR, Dubna
• 2Fermilab

#### Send email to authors

Public note CDF 11072

We present a top-quark mass measurement with dilepton events using the full CDF Run II data set. 520 events were obtained after applying the selection cuts for the data sample with the integrated luminosity of 9.1 $$\fbm$$. For our analysis we have built a variable that is expected to achieve the smallest uncertainty of the measurement. To determine this variable, we start from two initial observables. The first observable is the reconstructed top-quark mass which is calculated using all available experimental information and has the best sensitivity to the top-quark mass. To reduce the systematic uncertainty we consider a second observable that is insensitive to the JES. For the template analysis we use variable obtained as a weighted sum of the two above-mentioned variables. The weights are chosen to be optimal for our measurement. Templates are built from simulated $$\ttbar$$ and background events, and parameterized in order to provide probability distribution functions. The top-quark mass is extracted by performing a likelihood fit in data to a weighted sum of signal and background. We measure a top-quark mass value of $$171.46 \pm 1.91$$(stat)$$\pm 2.51$$(syst) $$\gevcc$$ or $$171.46\pm 3.15$$ $$\gevcc$$.

## Result

Mtop = 171.46 ± 1.91(stat) ± 2.51(syst) GeV/c2

### Tables and Plots

Description of Plots gif eps
1 Likelihood fit to the dilepton data sample. Background (purple) and signal+background (cyan) p.d.f.'s, normalized according to the numbers returned by the fitter, are superimposed to the $$\MTeff$$ distribution from data (points). Plot is for b-tagged subsample. eps
2 Likelihood fit to the dilepton data sample. Background (purple) and signal+background (cyan) p.d.f.'s, normalized according to the numbers returned by the fitter, are superimposed to the $$\MTeff$$ distribution from data (points). Plot is for non-tagged subsample. eps
3 The fitted mass-dependent negative log-likelihood function from the likelihood fit to the dilepton data sample. eps
4 Background (purple) and signal+background (cyan) templates for reconstructed top quark mass, normalized according to the numbers returned by the fitter, are superimposed to the reconstructed mass distribution from data (points). Plot is for b-tagged subsample. The value 171 $$\gevcc$$ is assumed for the top quark mass (close to the central value of the data fit). eps
5 Background (purple) and signal+background (cyan) templates for reconstructed top quark mass, normalized according to the numbers returned by the fitter, are superimposed to the reconstructed mass distribution from data (points). Plot is for non-tagged subsample. The value 171 $$\gevcc$$ is assumed for the top quark mass (close to the central value of the data fit). eps
6 Background (purple) and signal+background (cyan) templates for $$M_{lb}^{alt}$$ variable, normalized according to the numbers returned by the fitter, are superimposed to the reconstructed mass distribution from data (points). Plot is for b-tagged subsample. The value 171 $$\gevcc$$ is assumed for the top quark mass (close to the central value of the data fit). eps
7 Background (purple) and signal+background (cyan) templates for $$M_{lb}^{alt}$$ variable, normalized according to the numbers returned by the fitter, are superimposed to the reconstructed mass distribution from data (points). Plot is for non-tagged subsample. The value 171 $$\gevcc$$ is assumed for the top quark mass (close to the central value of the data fit). eps
8 Summary of uncertainties on the top quark measurement. eps
9 Summary table of expected contributions and observed events in SecVtx b-tagged and non-tagged dilepton data samples. eps
10 Uncertainties in the measured top quark mass as a function of w. eps