Empirical type 1 error rate for large sample LRT {solid line (full=black, partial= gray)}, score/log-rank test (dot/dashed line), and Wald test {dashed line (full=black, partial=gray)} as a function of sample size n = N (i.e., N/2 per arm) for p0 = p1 = 0.5 probability of infection per challenge under Scenario 1 (no heterogeneity or immunity) and various values of cmax = Cmax. The dotted horizontal line corresponds to the nominal significance level α = 0.05.