Versión de HTML Básico
Assessing the Effect of Conditional Cash Transfers in Children Chronic Stunting: The Human Development Bonus in Ecuador
111
13
Analiti a, Revista de análisis estadístico, Vol. 13 (1), 2017
Table 9:
Intention-to-treat (non-parametric)
VARIABLES
(1)
(2)
(3)
Conventional
-0,286* -0,330 -0,434**
(0,165) (0,214) (0,200)
Bias-corrected
-0,304* -0,348 -0,484**
(0,165) (0,214) (0,200)
Robust
-0,304 -0,348 -0,484**
(0,202) (0,240) (0,227)
Controls
NO NO YES
Observations
6,321 6,321 6,321
Obs. left main-bandwidth
549
637
627
Obs. right main-bandwidth 593
713
704
Obs. left bias-bandwidth
700
803
768
Obs. right bias-bandwidth
816
973
897
Conventional p-value
0,0834 0,123 0,0297
Robust p-value
0,133 0,147 0,0327
Order Loc. Poly. (p)
1
2
2
Order Bias (q)
2
3
3
BW Loc. Poly. (h)
5,252 6,338 6,248
BW Bias (b)
7,272 8,553 8,015
Standard errors in parentheses
*** p
<
0,01, ** p
<
0,05, * p
<
0,1
Dep var= HAZ (Y)
Note:
(1), (2) and (3) present non-parametric estimates of the effect of the eligibility
status on stunting z-scores. The table shows conventional, bias-corrected and robust
estimates, and the specificities of the sample since the bandwidths are data-driven.
Statistical significance is not homogeneous, though they are all negative. For instance,
the significant quadratic specification robust estimate can be interpreted as a decrease
of 0,48 sd. in HAZ due to an eligible index for households in the vicinity (1.331
observations)
29