= Solution
For a selected <lepton> sample, subtract backgrounds and correct for <detector acceptance>, reconstruction and particle-identification efficiency, trigger efficiency and live time. Schematically,
$$
\boxed{\sigma_{\rm fid}=\frac{N_{\rm obs}-N_{\rm bg}}{\mathcal L_{\rm int}\,\epsilon},\qquad
\sigma_{\rm total}=\frac{N_{\rm obs}-N_{\rm bg}}{\mathcal L_{\rm int}\,A\epsilon}.}
$$
The first expression applies when the published signal definition is a specified fiducial region; the second additionally extrapolates by acceptance $A$. Resolution-induced migration across selections must be unfolded or included in a detector-response fit. <Tau lepton> reconstruction also needs the relevant <branching fractions> if only selected decay modes are counted. Initial- and final-state <QED> radiation and beam-<energy> spread change the measured line shape; extracting a Born-level or pole <cross section> requires the corresponding radiative corrections. In the <electron> channel, <Bhabha scattering> also has a non-annihilation $t$-channel contribution that must be treated.
For <quarks>, <hadronization> means the detector observes hadrons and <jets>, rather than free <quarks>. Correct the selection and any flavor tagging using fragmentation and decay modeling, including losses to heavy-hadron decays with <neutrinos>. In comparison with a pointlike <lepton> <cross section>, also include the three <color charges> per <quark flavor> and <QCD> corrections, schematically $N_c[1+\alpha_s/\pi+\cdots]$ with $N_c=3$ at high <energy>. The color multiplicity is a physical production factor, not an extra factor by which to divide an already efficiency-corrected inclusive hadronic event count.
Back to article page
