(A-B). Uncertainty in the incubation period distribution was accounted for by updating the parameters, and , of a lognormal incubation period distribution (these parameters represent the mean and standard deviation of the natural logarithm of the incubation period, respectively) alongside unknown model parameters during the MCMC procedure. Independent normal prior distributions (truncated at zero) consistent with the 95% confidence intervals obtained by McAloon et al., 2020 were assumed for (prior mean 1.63 log(day), standard deviation 0.061, 95% CrI 1.51–1.75; panel A) and (prior mean 0.5 log(day), standard deviation 0.026, 95% CrI 0.45–0.55; panel B). This incubation period was used directly when evaluating the likelihood in the independent transmission and symptoms model. In the mechanistic model, we assumed a gamma distributed incubation period with the same mean and standard deviation as a lognormal distribution with parameters and . (C-F). Violin plots indicating posterior distributions of the mean (C) and standard deviation (D) of the generation time distribution, proportion of transmissions occurring prior to symptom onset (among infectors who develop symptoms; E), and overall infectiousness parameter, (F), for the independent transmission and symptoms model with either a fixed (as in Figure 1; blue) or variable (i.e. accounting for uncertainty in and as described above; red) incubation period distribution, and for the mechanistic model with a fixed (orange) or variable (purple) incubation period distribution.