Skip to main content
. Author manuscript; available in PMC: 2015 Dec 1.
Published in final edited form as: Health Serv Outcomes Res Methodol. 2014 Aug 7;14(4):166–182. doi: 10.1007/s10742-014-0123-z

Table III.

Bias and Nominal 95% Confidence Interval Coverage rates of treatment effect estimation using 6 DD models

Setting Bias

(1): Naïve
DD
(2): DD
with X
(3): (1)
with
separate
wts
(4): (2)
with
separate
wts
(5): (1)
with 4
group
weighting
(6): (2)
with 4
group
weighting
1 −0.0047 −0.0048 −0.0047 −0.0047 −0.0048 −0.0048

2 0.060 0.060 −0.023 −0.023 −0.00043 −0.00064

3 0.112 0.111 0.029 0.029 0.0017 0.0018

4 0.204 0.089 −0.034 −0.035 0.00026 0.00024

Confidence interval coverage
(Nominal=.95)

(1): Naïve
DD
(2): DD
with X
(3): (1)
with
separate
wts
(4): (2)
with
separate
wts
(5): (1)
with 4
group
weighting
(6): (2)
with 4
group
weighting
1 .94 .95 .97 .95 .96 .95

2 .91 .90 .96 .94 .96 .94

3 .79 .76 .95 .93 .96 .95

4 .44 .83 .94 .92 .95 .93