Skip to main content

Table 3 Goodness of fit when fitting the EME and the HMD datasets to the four-parameter model and the Siler model

From: A two-parameter hazard function to describe age patterns of mortality in ancient Northwestern Europe

  Mean WRMSEa
Dataset 4-Parameter modelb
(SD)d
Silerc
Model (SD)d
Per period   
EME 1640–1699 (n = 6) 0.1620 (0.0510) 0.1912 (0.0566)
EME 1700–1749 (n = 5) 0.1623 (0.0330) 0.1695 (0.0458)
EME 1750–1809 (n = 6) 0.1467 (0.0522) 0.1631 (0.0515)
HMD 1750–1800 (n = 50) 0.1025 (0.0250) 0.1126 (0.0278)
HMD 1801–1850 (n = 127) 0.0706 (0.0169) 0.0767 (0.0214)
HMD 1851–1900 (n = 350) 0.0598 (0.0216) 0.0626 (0.0300)
HMD 1901–1950 (n = 345) 0.0363 (0.0492) 0.0494 (0.0613)
HMD 1951–2000 (n = 350) 0.0259 (0.0104) 0.0214 (0.0074)
HMD 2001–2015 (n = 98) 0.0397 (0.0109) 0.0294 (0.0120)
Per country   
HMD Belgium 1841–2015 0.0403 (0.0209) 0.0393 (0.0195)
HMD Denmark 1835–2014 0.0420 (0.0210) 0.0353 (0.0195)
EME England 1640–1809 0.1567 (0.0475) 0.1749 (0.0532)
HMD England & Wales 1841–2013 0.0399 (0.0334) 0.0495 (0.0519)
HMD France 1816–2014 0.0582 (0.0555) 0.0714 (0.0707)
HMD Netherlands 1850–2014 0.0356 (0.0216) 0.0351 (0.0215)
HMD Norway 1846–2014 0.0448 (0.0297) 0.0465 (0.0327)
HMD Sweden 1751–2014 0.0537 (0.0354) 0.0578 (0.0367)
All 1337 life tables 0.0473 (0.0371) 0.0506 (0.0451)
  1. Notes:
  2. aWRMSE weighted root mean square error (see equation 6). The smaller the error the better the fit.
  3. bSee equations 25.
  4. cSee equation 1.
  5. dSD standard deviation
  6. Sources data: Human Mortality Database (2017, 2017, 2018); Wrigley, Davies, Oeppen & Schofield (1997: 224, 239, 250, 290)