Evaluation of goodness of fit (change in scaled deviance) of an age-period-cohort modelling technique for number of deaths from pleural mesothelioma, 1969–98
| Men | Women | ||||||
|---|---|---|---|---|---|---|---|
| Terms in model | Reference for significance testing | Scaled deviance | df | p value | Scaled deviance | df | p value |
| Intercept (1) | – | 9190 | 359 | 1397 | 359 | ||
| Age (2) | 1 | 1081 | 348 | 0.0000 | 338 | 348 | 0.0000 |
| Cohort (3) | 1 | 4190 | 346 | 0.0000 | 599 | 346 | 0.0000 |
| Period (4) | 1 | 8563 | 354 | 0.0000 | 1375 | 354 | 0.0007 |
| Age + cohort (5) | 2 | 335 | 335 | 0.0000 | 313 | 335 | 0.0211 |
| Cohort + period (6) | 3 | 1156 | 341 | 0.0000 | 350 | 341 | 0.0000 |
| Age + period (7) | 2 | 406 | 343 | 0.0000 | 318 | 343 | 0.0011 |
| Age + period + cohort (8) | 5 | 315 | 330 | 0.0013 | 301 | 330 | 0.0394 |
| APC two-period, cut off 1996 (9) | 5 | 311 | 334 | 0.0000 | 303 | 334 | 0.0012 |