Skip to main content

Table 4 Validation of the predictive models

From: Public health application of predictive modeling: an example from farm vehicle crashes

Validation data years (Training data years) Total validation data occupants (training data occupant) Observed injuries or deaths in validation data (N) Non-modifiable (Model 1) Non + semi-modifiable
(Model 2)
Non + semi + modifiable
(Model 3)
Avg. pred. Prob. Expected injuries/ deaths (N) Avg. pred. Prob. Expected injuries/ deaths (N) Avg. pred. Prob. Expected injuries/ deaths (N)
2008-‘10 (‘05-‘07) 7624 (7210) 1095 0.1407b 1073 0.1445e 1102 0.1437h 1095
2009-‘10 (‘05-‘08) 5216 (9618) 779 0.1383c 721a 0.1405f 733 0.1397i 729
2010 (‘05-‘09) 2615 (12,219) 372 0.1424d 372 0.1410g 369 0.1401j 366
  1. Abbreviations: Avg. pred. Prob. Average predicted probability
  2. ap-value = 0.0327 (Chi Sq = 4.56, df = 1), suggesting that expected injuries and deaths (n = 721) were significantly different than the observed (n = 779). All other expected to observed differences were non-significant
  3. bAUC = 0.69 based on training data 2005–2007
  4. cAUC = 0.70 based on training data 2005–2008
  5. dAUC = 0.69 based on training data 2005–2009
  6. eAUC = 0.75 based on training data 2005–2007
  7. fAUC = 0.75 based on training data 2005–2008
  8. gAUC = 0.74 based on training data 2005–2009
  9. hAUC = 0.77 based on training data 2005–2007
  10. iAUC = 0.77 based on training data 2005–2008
  11. jAUC = 0.77 based on training data 2005–2009