Page 35 - IJPS-2-2

Mukesh Ranjan, Laxmi Kant Dwivedi, Rahul Mishra, et al.

social, economic, and environmental factors.

2. Data and Methods

2.1 Sources of Data
This study used data from the third round of the District Level Household Survey (DLHS-III) con-
ducted by the Government of India during 2007–2008. It utilized pregnancy history files for ever
married women (aged 15–49 years) including questions related to the outcome of their pregnancy
and the birth and death of each of their children born since January 2004 until the date of survey. The
pregnancy history file was also used to compute infant deaths. We did not take multiple births of a
pregnancy of a mother for analysis since the multiple births are relatively rare. The detailed sampling
procedure and data quality were available in the official document released by the International In-
stitute for Population Sciences (2010).
2.2 Measurements

2.2.1 Tribes and Non-Tribes
The caste variable was recoded into two categories: scheduled tribes (STs) and non-tribes. The
non-tribes category included scheduled caste (SC), other backward classes, and others classes (i.e.,
other than SC/ST/OBC classes). The entire analysis was based on these two categories only in order
to estimate mortality differentials between tribes and non-tribes.
2.2.2 Other Factors Associated with Infant Mortality

To identify important factors affecting infant and child mortality among the tribal and non-tribal
populations of Central and Eastern parts of India, a set of possible factors was considered but only
variables that met either of the following two criteria were considered in the analyses. Firstly, there
is at least a moderate bivariate association (p < 0.05) between the selected factor and infant mortality
for at least one out of four states studied. Secondly, in some cases, importance was given to the
theoretical rather than purely statistical association. The theoretical association refers to the inclusion
of factors in the present study which prior literature has evidenced to be associated with infant mor-
tality. For example, at the individual level, the risk of death of a child is influenced to a great ex-
tent by factors related to the mother: her education; her situation prior to and post pregnancy; care
received before, during, and after pregnancy; location of birth; birth order; and care received by
the child during the first few years of his or her life. Apart from that, the anthropological literature
has also focused on access to health care for India’s adivasi. The evidence from the grassroots high-
lights distinct problems in tribal areas because of higher poverty, poorer health, and lower education
as compared to non-tribal areas. There is a wide acknowledgment that excessive childhood mortality
in tribes is partly due to poverty and partly due to poor access to services.
Overall, we used the following variables in the model which are categorized as follows. Moth-
er-specific variables include mother’s age, mother’s education, and feeding colostrum. In-
fant-specific variables include birth order and sex of the infant. Household/community-specific va-
riables include state of residence, place of residence, religion, and the wealth index. Program varia-
ble includes place of a pregnancy registration.
Mother’s age was classified into three categories: 15–24 years, 25–34 years, and 35+ years. Many
studies have suggested that the age of the mother is a strong predictor affecting the survival chances
of an infant. Hence, we categorized age of the mother into categories in such a manner so that the
risk attributable to mother’s age gets captured properly. Mother’s highest years of schooling
was classified into four categories: illiterate, primary school, secondary school, and high school or
above. The question of mother’s highest years of schooling was only asked by those women who
responded “yes” to the question of whether the woman ever attended a school. In this case we found

International Journal of Population Studies | 2016, Volume 2, Issue 2 29