Skip to main content
. 2020 Apr 28;24(8):1028–1037. doi: 10.1007/s10995-020-02943-1

Table 3.

Determinants of the probability that child is found from the health post records (marginal effects from probit regressions)

(1)
Number of observations (1 child per household): 2,155
Dependent variable: Child is found from the health post records
Master Family Index exists at the health post 0.032
(0.074)
Family Folders exist at the health post 0.208***
(0.076)
Household's distance (in km) to the health post − 0.032***
(0.012)
Number of years the HEW has worked in the kebele 0.002
(0.006)
Child's age in months − 0.004**
(0.002)
Child is male 0.019
(0.020)
Child was born home − 0.040
(0.030)
Mother's age (in years) 0.001
(0.002)
Mother's education (in years) 0.003
(0.005)
Household size − 0.010
(0.007)
Household has other children who are less than 5 years of age 0.051*
(0.026)
Number of tropical livestock units (TLU) owned 0.004
(0.004)
Household benefitted from PSNP in 2016 0.018
(0.026)
Amhara region − 0.089
(0.056)
Oromia region − 0.195***
(0.065)
SNNP region − 0.106*
(0.059)
Tigray region (reference)

The non-italicized rows indicate variables, coefficients and standard errors. The italicized rows provide additional information about the model; dependent variable and the sample size

Reported coefficients are marginal effects from a probit model. Standard errors are reported in parentheses, and are clustered at the health post (kebele) level. The standard errors for the marginal effects are calculated using the Delta Method

*, ** and *** indicate significance at the 10%, 5% and 1% level, respectively