Through a counterexample, Redinbo’s upper bound on the ratio of the probability of a group and that of its coset is shown to be incorrect. The given example is also shown to provide a new upper bound. © 1981 IEEE
