Page 133 - IJPS-2-2

Priyanka Dixit and Laxmi Kant Dwivedi

All models used a logit link function to obtain estimates of fixed and random effects and are esti-
mated with MLwiN software (version 2.15, http://www.bristol.ac.uk/cmm/software/mlwin/). For
interpretive ease, fixed coefficients were converted to odds ratios and the significance level of p <
0.05 was used.

3 Results
3.1 Percentage Using Institutional Delivery Care

Table 1 presents the percent distribution of ever-married women according to the utilization of insti-
tutional delivery for up to three of the most recent (successive) births in India in the five years prior
to the NFHS-III in 2005-2006. The distributions show that the prevalence of utilization of institu-
tional delivery among women who had one, two, and three births in the five years prior to the survey
was 47.4%, 26.2%, and 15.6%, respectively. Among women who had three births, 63.1%of them
delivered all three births at home and 5% delivered their third birth in a hospital while the previous
two births were at home.
3.2 Validation of Assumption of Independent Bernoulli Trials

We first applied a three-level model to examine the consistent use of institutional delivery (presented
2
in Table S1 in the Appendix). In the three-level model keeping σ as unconstrained, we validated
e
2
the assumption of independent Bernoulli trials. Our estimated σ value is 0.16 (see Table S2),
e
which was far below 1 and highlights the problem of under-dispersion in the model. This clearly
shows that pregnancies within mothers are not conditionally independent of each other — note:

Table 1. Percent distribution of ever-married women according to the consistent utilization of institutional delivery for successive births in India.
Observed births in sequence of increasing birth order Consistent utilization of institutional delivery
% N

Women With One Birth

Home 52.56 10956
Institute 47.44 13080
Total 100 24036
Women With Two Births

Home, Home 59.73 6007
Home, Institution 5.87 686
Institution, Home 8.21 890
Institution, Institution 26.19 3448
Total 100 11031
Women With Three Births

Home, Home, Home 63.05 1006
Institution, Home, Home 5.98 95
Institution, Institution, Home 2.98 48
Institution, Institution, Institution 15.55 301
Home, Institution, Institution 2.60 58
Home, Home, Institution 5.08 96
Home, Institution, Home 2.24 31
Institution, Home, Institution 2.51 43
Total 100 1678
Note: Successive births begin with the first birth for which information is available in the 5-year period before the NHFS-III.

International Journal of Population Studies | 2016, Volume 2, Issue 2 127