The scientific rationale for using live animals in car crash studies proceeds from the argument that comparative biomedical and biomechanical data are needed to develop an instrumented dummy, or anthropomorphic test device, which will provide reliable, reproducible information for designing safe cars. The animal studies are thus not really ends in themselves, i.e., they do not supply data which can be readily applied to real situations. Instead, they contribute to a pool of information which is supposed to lead to the perfecting of an experimental subject (the instrumented dummy) which will eventually render the further use of Iive animals unnecessary.
Heneson, N. (1980). Live Animals in Car Crash Studies. International Journal for the Study of Animal Problems, 1(4), 214-217.