Skip to main content
. 2017 Sep 29;7:34. doi: 10.1186/s13561-017-0169-z

Table 4.

Estimation results: Probit, Probit with Interaction and PSM

Dependent Var: Neonatal Deaths Model1 Model 2 Model 3
Probit Probit with Interaction PSM
Woman has Valid NHIS −0.0144* −0.0646***
(0.00795) (0.0246)
Mother’s Age −0.000490 −0.000528
(0.000586) (0.000589)
Child is Twin 0.000431 0.000148
(0.0146) (0.0147)
Birth Order 0.00282 0.00295
(0.00194) (0.00194)
Child is Male 7.20e-05 −3.80e-06
(0.00492) (0.00492)
Number of ANC Visits 9.29e-05 0.000106
(0.000970) (0.000971)
Delivered by CS 0.00497 0.00381
(0.00729) (0.00737)
Birth Interval −0.00933 −0.00994
(0.00808) (0.00800)
Facility Delivery 0.00961 0.00942
(0.00731) (0.00727)
Distance to Health Facility a Problem 0.0107* 0.0117**
(0.00552) (0.00554)
Employed −0.00617 −0.00520
(0.00603) (0.00605)
(0.0175) (0.0177)
Married −0.0137 −0.0141
(0.00863) (0.00864)
Living Together −0.0224** −0.0227**
(0.0103) (0.0103)
Divorced/Separated/Widowed −0.0188 −0.0189
(0.0141) (0.0141)
Urban −0.000404 −0.0286*
(0.00678) (0.0150)
Urban*Valid NHIS 0.0323**
(0.0154)
Education Controls Yes Yes
Regional Controls Yes Yes
Wealth Quintiles Yes Yes
Valid NHIS vs No Valid NHIS −0.051***
(0.007)
Observations 2905 2905 2905

Standard errors in parentheses*** p < 0.01, ** p < 0.05, * p < 0.1

Marginal effects presented in Models 1 and 2