Lorena Moreno
112
Analiti a, Revista de análisis estadístico, Vol. 13 (1), 2017
Figure 8:
Intention-to-treat (non-parametric robust)
Note:
ITT (1), ITT (2) and ITT (3) plot the robust estimates of the effect
of the eligibility status on stunting z-scores, given the different specifications.
All evidence lower HAZ values for households reporting index scores to the
left of the cutoff
30