Q: Do you trust the gamma conversion function extracted from Pythia Monte Carlo?

A:  Yes and No. We derive gamma conversion function from fully detector simulation Monte Carlo and scaled up factor of 1.3 from the data-driven study of gamma conversion rate measurement.

Q: there is no integration over the parton densities. Is this already included in the definition of dsigma(th)/dy ?

A:  Yes. The Parton densities is included in the definition of dsigma(th)/dy

Q: Is dy actually an integration over n dimension, what is the value of n ?

A:  Yes. dy is an integration over n dimension. The n equals 8 in the case of evaluating WW/HWW/ZZ event probability and it equals 5 for Wgamma and Wjet.

Q: How does lepton flavour come in ? Pm(xobs) is a per event probability. Each event has two leptons. Does this mean that sigma(theo) is different for every channel, ee, emu, mumu ?

A:  The sigma(theo) is the same to ee, emu, mumu. The matrix element from MCFM treated final state particle as massless paricles. We rescale the observed energy of muon to massless particle before calculating the sigma(theo).

Q: In the text it then says that epsilon(y) is the probability to reconstruct a parton level object as one of the seven lepton categories. I don't understand how this work, since it is not a per event efficiency. Is it different for every category, ee, emu etc. ?

A:  In CDF, we categorize leptons into seven types (2 electron, 4 muon and 1 track) and four of them are triggerable leptons. The number of dilepton category is 22. We use Monte Carlo to get the reconstruction efficiency, epsilon(y).