. Author manuscript; available in PMC: 2014 Sep 10.

Published in final edited form as: Stat Med. 2013 Apr 1;32(20):3552–3568. doi: 10.1002/sim.5802

Table 3.

Estimates of the PATE and generalizability bias due to age exclusion in studies of the effect of assisted delivery instrument choice on severe laceration. The second and third column from the left display the estimates as described in Section 3.2, and the rightmost two columns display the estimates where SPATE is estimated from the data, but the proportion of the excluded population is artificially set to 0.75.

		${\hat{π}}_{E} = 0.05$		π_E = 0.75
	$\hat{Δ}$	$\hat{γ}$	% Bias	$\hat{γ}$	% Bias
Simple	−0.13042 (−0.14012, −0.12073)	−0.00169 (−0.00390, 0.00052)	1.297	−0.02511 (−0.05786, 0.00764)	19.250
IPW	−0.12805 (−0.13852, −0.11758)	−0.00259 (−0.00491, −0.00026)	2.021	−0.03842 (−0.07287, −0.00396)	30.000
DR	−0.12711 (−0.13758, −0.11664)	−0.00237 (−0.00469, −0.00004)	1.863	−0.03515 (−0.06964, −0.00066)	27.655