Table 3 Threshold models for influenza population loss with Calcutta outlier
From: A geographic analysis of population density thresholds in the influenza pandemic of 1918–19
Model specification: District-specific intercepts and growth rates | |||
|---|---|---|---|
Threshold point estimate | Alternate threshold 1 | Alternate threshold 2 | |
Time trend (β10) | † | † | † |
Flu dummy (β20) | −0.7676*** | −0.2997*** | −0.3201*** |
(0.1546) | (0.0128) | (0.0174) | |
Low density * flu dummy (β30) | 0.4894*** | 0.0726*** | 0.0654*** |
(0.1550) | (0.0250) | (0.0224) | |
Time trend * flu dummy (β40) | † | † | † |
Number of obs. | 1194 | 1194 | 1194 |
R 2 | 0.9963 | 0.9963 | 0.9963 |
KEY DEMOGRAPHIC PHENOMENA | |||
Threshold population density | 19067 | 175 | 435 |
Range of possible thresholds (5% level of significance) | 1138-19067 | 175-207 | 430-464 |
Number (percentage) of districts outside threshold range | 52 (26%) | ||
Population loss as % of population, low density districts | −4.44% | −3.72% | −3.51% |
Population loss as % of population, high density districts | −21.15% | −4.81% | −6.06% |