Table 2 Details of parameter estimates with standard errors (s.e.) for ten different settings for COVID-19 cluster size data [8]. Values are given for the dispersion parameter, or k-parameter, defined for all models fitted and the c-parameter for the EPPM models. A value of 0 for the c-parameter corresponds to a Poisson model and a value of 1 a negative binomial model
From: Comparisons of statistical distributions for cluster sizes in a developing pandemic
Setting | Mean estimate (s.e.) | k-parameter estimate (s.e.) | c-parameter estimate from EPPM formulation (s.e.) |
|---|---|---|---|
Elderly care | 37.95 (9.01) | 0.86 (0.11) | 1 |
Food processing plants | 187.47 (44.10) | 0.86 (0.11) | 1 |
Household | 3.68 (0.60) | ∞ | 0 |
Large shared accommodation | 85.19 (37.23) | 0.23 (0.29) | 1.24 (0.13) |
Meal | 6.03 (1.53) | 0.64 (0.34) | 3.97 (1.19) |
Party | 29.14 (8.50) | 0.86 (0.11) | 1 |
Religious | 39.77 (9.22) | 0.86 (0.11) | 1 |
School | 33.91 (11.13) | 0.86 (0.11) | 1 |
Sport | 5.86 (1.44) | 0.86 (0.11) | 1 |
Work | 11.62 (3.04) | 0.42 (0.24) | 3.34 (0.76) |