Test of the <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"> <mml:mrow> <mml:mi mathsize="small">G</mml:mi> <mml:mi mathsize="small">E</mml:mi> <mml:mi mathsize="small">N</mml:mi> <mml:mi mathsize="small">I</mml:mi> <mml:mi

The reduced cross section of the semiexclusive <a:math xmlns:a="http://www.w3.org/1998/Math/MathML" display="inline"> <a:mo stretchy="false">(</a:mo> <a:mi>l</a:mi> <a:mo>,</a:mo> <a:msup> <a:mi>l</a:mi> <a:mo>′</a:mo> </a:msup> <a:mi>p</a:mi> <a:mo stretchy="false">)</a:mo> </a:math> lepton scattering process irrespective of the type of interaction is determined mainly by the bound nucleon momentum distribution in the target and nucleon final state interaction with the residual nucleus. These cross sections can be identified with distorted nuclear spectral functions and therefore, they are similar up to Coulomb corrections for neutrino and electron scattering on nuclei. In this article, we exploit this similarity and use data with precise kinematics and large statistics for semiexclusive electron scattering on an oxygen target to test models employed in the neutrino event generator. We find that these models cannot reproduce well the measured reduced cross sections in all allowed kinematic regions and the event generator needs to describe better both the nuclear ground states and nucleon final state interaction. The approach presented in this paper provides a great opportunity to better test the accuracy of nuclear models of quasielastic neutrino-nucleus scattering, employed in neutrino event generators.