CDF notes on
statistics issues
This list is obviously not exhaustive. Please let us know
of any good notes that we missed.
Some of the earlier CDF notes are
not linked to the database; only paper copies of these
exist.
The Statistics Committee considers that the CDF notes
listed below contain interesting statistical analyses. We
would not, however necessarily endorse every aspect of each of
these analyses. In case of doubt, please contact the
Statistics Committee.
-
Statistical
tests
- H.Keutelian,
"The Kolmogorov-Smirnov statistic when estimating
parameters from data"
[Describes a simple Monte Carlo procedure to convert a KS
statistic into a probability when comparing a data sample
to a model in which one or more parameters have been
determined from the data.]
CDF note 1285, Version 1.0, April 30,
1991.
- L. Demortier,
"Assessing the significance of a deviation in the tail of a
distribution"
[Defines and investigates some properties of the
Kolmogorov-Smirnov, Smirnov-Cramer-von Mises and
Anderson-Darling statistics for one-sample and two-sample
tests.]
CDF note 3419, Version 1.0, November 20, 1995.
- A. Beretvas, M. Binkley, L.
Demortier,
"Binned KS probability"
[Studies the effect that binning a data sample has on the
conversion of a KS statistic into a probability. Especially
useful to interpret the results of the PAW command
"diff".]
CDF note 4373, Version 2.0, November 25, 1997.
- T. Dorigo,
"Kolmogorov Tests of the Superjet Events"
CDF note 4861
- T. Dorigo,
"Study of the Significance of the Dimuon Mass Bump and the
Greedy Bump Bias"
CDF note 5239
- L. Lyons,
"Selecting Between Two Hypotheses"
OUNP-99-12, August 1999.
-
Fitting
techniques
- T. Devlin,
"Correlations from systematic corrections to
Poisson-distributed data in log-likelihood functions"
[Describes how to incorporate the effect of systematic
uncertainties, background contamination and efficiency
corrections in Poisson likelihood fits.]
CDF note 3126, version 3.1, May 27, 1999.
- H. Keutelian,
"A procedure to apply multiple mass constraints to a set of
tracks"
CDF note 3565, February 21, 1996.
- F. Bedeschi,
"A technique to break-up statistical and systematic errors"
CDF note 6859, January 27, 2004.
- Jonas Rademacker,
"Reduction of Statistical Power Per Event Due to
Upper Lifetime Cuts in Lifetime Measurements"
CDF note 7359, November 9, 2004.
- Luc Demortier,
"Equivalence of the best-fit and covariance-matrix
methods for comparing binned data with a model in the
presence of correlated systematic uncertainties"
CDF note 8661, April 28, 1999.
-
Unfolding
methods
- A. Robinson, P.K. Sinervo,
"Correcting the top quark pT distribution"
[This analysis uses an unfolding method that's designed to
work with small numbers of events and with background.]
CDF note 4821, December 17, 1998.
- A. Bhatti, B. Flaugher,
"Inclusive jet cross section in pbar-p collisions at
sqrt(s)=1.8 TeV"
[Section 5 of this PRD describes the unfolding method. It
is designed to work with large numbers of events and no
background.]
CDF note 5291, version 4.0, June 24, 2000.
-
Combining
measurements
- R.J. Hollebeek, P.K. Sinervo, H.H. Williams,
"The evaluation of upper limits for top quark production
using combined measurements"
[Defines upper limits according to classical and Bayesian
statistics, and descibes three methods to obtain a combined
upper limit from two or more measurements. Also describes
how to incorporate systematic uncertainties in the
calculation of upper limits. Although the authors clearly
favor the frequentist approach (often referring to it as
"anti-Bayesian"), their method of smearing with systematics
is Bayesian.]
CDF note 1109, January 6, 1990.
- M. Shochet,
"More on statistics of combined measurements"
[Studies the three methods described in CDF note 1109 and
concludes that only one of them is robust.]
CDF note 1212, May 25, 1990.
- T. Liss, G. Unal, P.K.
Sinervo,
"Estimating the statistical significance of a top quark
search counting experiment."
[Calculates the significance of the Run 1A top quark
sample, by combining the events observed in the dilepton
and tagged lepton+jets channels.]
CDF note 2370, Version 1.1, December 1, 1993.
- L. Demortier, R. Hughes, R.
Roser,
"Averaging the top quark mass measurements from CDF and
D0"
[The appendix describes the mathematical formalism used to
combine measurements. It is based on minimizing a chisquare
with a covariance matrix that includes statistical and
systematic uncertainties and their correlations.]
CDF note 4439, version 2.5, October 13, 1998.
- C. Grosso-Pilcher, G.
Landsberg,
"Combined limits on first generation leptoquarks from the
CDF and D0 experiments"
[Does both a frequentist and Bayesian analysis.]
CDF note 4575, version 2.0, May 11, 1997.
- J. Conway,
"Calculation of Cross Section Upper Limits Combining
Channels Incorporating Correlated and Uncorrelated
Systematic Uncertainties",
CDF note 6428, April 23, 2003
-
Confidence intervals
and limits
- C. Blocker,
"Dilution factors"
[Discusses how the statistical uncertainty of a measurement
is affected by backgrounds, mistaggings and nonzero
resolutions. Applies mainly to asymmetry and mixing
measurements.]
CDF note 3406, Version 1.1, December 12, 1995.
- J. Conway, K. Maeshima,
"Upper limits on Poisson processes incorporating
uncertainties in acceptance and background"
CDF note 4476, Version 2, March 13, 1998.
- C. Blocker,
"Comment on the Feldman and Cousins method of limits"
[How to set limits near physical boundaries; applied to the
sin(2beta) analysis.]
CDF note 4611, Version 1.0, May 27, 1998.
- J. Heinrich,
"Coverage of Error Bars for Poisson Data"
CDF note 6438, May 2, 2003
-
Monte
Carlo methods
- M. Mangano,
"The unweighting of matrix element Monte Carlos"
CDF note 1665, January 1991.
-
Systematic
Uncertainties