I run the following four logit regressions:
The results are
I was expecting, in columns (2) and (4), the estimates of odds ratio, rather than the coefficients. So, any suggestions? Thanks.
Code:
webuse lbw, clear logit low age lwt i.race smoke, nolog estimates store m1a logit low age lwt i.race smoke, nolog or estimates store m1b logit low age lwt i.race smoke ptl ht ui, nolog estimates store m2a logit low age lwt i.race smoke ptl ht ui, nolog or estimates store m2b esttab m1a m1b m2a m2b, b(4) se(4) mtitle nogap
Code:
----------------------------------------------------------------------------
(1) (2) (3) (4)
m1a m1b m2a m2b
----------------------------------------------------------------------------
low
age -0.0225 -0.0225 -0.0271 -0.0271
(0.0342) (0.0342) (0.0365) (0.0365)
lwt -0.0125 -0.0125 -0.0152* -0.0152*
(0.0064) (0.0064) (0.0069) (0.0069)
1.race 0.0000 0.0000 0.0000 0.0000
(.) (.) (.) (.)
2.race 1.2312* 1.2312* 1.2626* 1.2626*
(0.5171) (0.5171) (0.5264) (0.5264)
3.race 0.9436* 0.9436* 0.8621* 0.8621*
(0.4162) (0.4162) (0.4392) (0.4392)
smoke 1.0544** 1.0544** 0.9233* 0.9233*
(0.3800) (0.3800) (0.4008) (0.4008)
ptl 0.5418 0.5418
(0.3462) (0.3462)
ht 1.8325** 1.8325**
(0.6916) (0.6916)
ui 0.7585 0.7585
(0.4594) (0.4594)
_cons 0.3301 0.3301 0.4612 0.4612
(1.1076) (1.1076) (1.2046) (1.2046)
----------------------------------------------------------------------------
N 189 189 189 189
----------------------------------------------------------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001
Comment