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

A:  We derive the gamma conversion function from fully detector-simulated Monte Carlo with a scaling factor of 1.3 from the data-driven study of the 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 are included in the definition of dsigma(th)/dy

Q: Is dy actually an integration over n dimensions? What is the value of n?

A:  Yes. dy is an integration over n dimensions. n equals 8 in the case of evaluating the 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 for ee, emu, mumu. The matrix element from MCFM treated final state particles as massless particles. We rescale the observed energy of a muon to a 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 works since it is not a per event efficiency. Is it different for every category, ee, emu etc. ?

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