Proportion of mice infected, mean incubation period and variance of incubation period at each relative dose.

At each relative dose the proportion infected (column 6) was estimated from all mice inoculated at that relative dose from 119 experiments for which an ID50 could be calculated. The mean (column 7) and variance (column 8) of the incubation period were estimated from all mice infected at that relative dose for which incubation period data were available (column 4). Part of the variability in incubation periods at each group is because of variability in the average incubation period between experiments. Column 9 shows the variance at each relative dose, excluding between-experiment variability in the average incubation periods. This metric was derived by first grouping mice according to experiment number and relative dose. For each mouse with an incubation period, we then calculated the difference between its incubation period and the mean incubation period of mice in that experiment-dose group. Data from experiment-dose groups for which there were fewer than two infected mice with an incubation period were excluded from this analysis (column 5). At each relative dose, the variance of these differences was then calculated (column 9).