Session 3 reading early development, parental practices, poverty and access to early education
Early childhood development: impact of national
human development, family poverty, parenting
practices and access to early childhood education
T. D. Tran,* S. Luchters† and J. Fisher*
*Jean Hailes Research Unit, School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC, Australia, and
†Centre for International Health, Burnet Institute, Melbourne, VIC, Australia
Accepted for publication 31 July 2016
Keywords
child development, low-
and middle-income
settings, poverty
Co
espondence:
Thach Duc Tran, Jean
Hailes Research Unit,
School of Public Health
and Preventive Medicine,
Monash University, Level
1, 549 St Kilda Road,
Melbourne, VIC 3004,
Australia
E-mail:
XXXXXXXXXX
Abstract
Background This study was to describe and quantify the relationships among family poverty,
parents’ caregiving practices, access to education and the development of children living in low-
and middle-income countries (LAMIC).
Methods We conducted a secondary analysis of data collected in UNICEF’s Multiple Indicato
Cluster Surveys (MICS). Early childhood development was assessed in four domains: language-
cognitive, physical, socio-emotional and approaches to learning. Countries were classified into three
groups on the basis of the Human Development Index (HDI).
Results Overall, data from XXXXXXXXXXchildren aged 36 to 59months from 35 LAMIC were included in
the after analyses. The mean child development scale score was 4.93 out of a maximum score of 10
(95%CI 4.90 to 4.97) in low-HDI countries and XXXXXXXXXX%CI 7.05 to 7.12) in high-HDI countries. Family
poverty was associated with lower child development scores in all countries. The total indirect effect
of family poverty on child development score via attending early childhood education, care for the
child at home and use of harsh punishments at home was !0.13 SD (77.8% of the total effect) in
low-HDI countries, !0.09 SD (23.8% of the total effect) in medium-HDI countries and !0.02 SD (6.9%
of the total effect) in high-HDI countries.
Conclusions Children in the most disadvantaged position in their societies and children living in
low-HDI countries are at the greatest risk of failing to reach their developmental potential.
Optimizing care for child development at home is essential to reduce the adverse effects of poverty
on children’s early development and subsequent life.
Introduction
Early childhood development, which is generally conceptual-
ized as comprising several domains, is a strong predictor of
adult health and productivity (Grantham-McGregor et al.
2007; Victora et al. 2008; Black & Hurley XXXXXXXXXXThe 2007 and
2011 Series on Child Development in The Lancet concluded
that more than 200 million children under the age of five years
fail to reach their development potential each year, and most of
them are living in resource-constrained settings (Grantham-
McGregor et al XXXXXXXXXXThe Series postulated the links between
poverty and inequalities in childhood development which are
mediated via biological factors including intrauterine growth
estriction, child undernutrition, micronutrient deficiencies,
infectious diseases and environmental exposures; and psycho-
social factors including early childhood education, parenting
practices and exposure to violence (Walker et al. 2007; Walke
et al XXXXXXXXXXHowever, there is as yet a lack of empirical
Original Article doi:10.1111/cch.12395
© 2016 John Wiley & Sons Ltd 415
s_bs_banner Child: care, health and development
evidence of the mechanisms of the effect of poverty on early
childhood development within and between countries.
The Multiple Indicator Cluster Surveys (MICS) (United
Nations Children’s Fund (Unicef)’s 2015) are household
surveys initiated by the United Nations Children’s Fund
(UNICEF) and implemented in up to five rounds in 108 low-
and middle-income countries. The MICS’ primary goal is to
monitor indicators of progress towards the Millennium
Development Goals related to women’s and children’s health
in these countries from the mid-1990s to 2015. Since the
fourth round (in 2010–2012), an early childhood development
indicator has been collected along with information about
caregiving practices for young children and household wealth.
Together, these provide unique data to examine the effects of
family poverty on early childhood development in diverse
settings.
The aim of this analysis is to assess the links among family
poverty, caregiving practices and early childhood development
using Multiple Indicator Cluster Survey Round Four data using
the model proposed in the 2011 Lancet Series on Child
Development that family poverty affects biological and
psychosocial factors, which in turn influence inequalities in
child development (Walker et al. 2011).
Methods
Study design and participants
In MICS Round Four (2010 – 2012), a large nationally
epresentative sample of between 5000 and XXXXXXXXXXhouseholds
was selected in each country using a multistage, cluster-
sampling technique. Early childhood development data were
collected for all children aged 36 to 59months in the selected
households. Only these children were included in this study.
Study measures
Child development as the main outcome measure was assessed
y a 10-binary fixed choice item scale (Bornstein et al. 2012)
encompassing four developmental domains including lan-
guage-cognitive (Can (name) identify or name at least ten letters
of the alphabet? Can (name) read at least four simple, popula
words? Does (name) know the name and recognize the symbol
of all numbers from 1 to 10?); physical (Can (name) pick up a
small object with two fingers, like a stick or a rock from the
ground? Is (name) sometimes too sick to play?); socio-
emotional (Does (name) get along well with other children?
Does (name) kick, bite or hit other children or adults? Does
(name) get distracted easily?); and approaches to learning (Does
(name) follow simple directions on how to do something
co
ectly? When given something to do, is (name) able to do it
independently?). These questions were derived from a
oad
set of indicators of child development developed by UNICEF
in 2007 and pilot-tested in Jordan, the Philippines and Kenya
(United Nations Children’s Fund (Unicef)’s XXXXXXXXXXEach item
is scored 1 if the child can achieve the task and 0 if they are not
able to. This yields a total score ranging from 0 (the least
optimal) to 10 (the most optimal) development.
Household economic status, our main exposure of interest,
was assessed using questions about household characteristics
including the main materials of the dwelling’s floor, roof and
exterior walls; main type(s) of fuel used for cooking; source of
drinking water; type of sanitation facility; and 12 durable
household assets. An index of household wealth was
constructed on the basis of these items using the World
Bank’s techniques for measuring living standards using
household survey data (O’Donnell et al. 2008).
Psychosocial factors including caregiving practices and early
childhood education consisted of information about (1)
whether the child was attending an organized early childhood
education programme; (2) whether in the past three days, the
mother or the father had engaged in any of six early learning
activities with the child including reading books; telling stores;
singing; naming, counting, and drawing; taking the child
outside or playing with the child; (3) the number of children’s
ooks in the household; (4) whether there were toys bought
from stores or manufactured available in the household; and
(5) harsh punishments were assessed for a randomly selected
child aged 2 to 14 years in the household. Five questions
adapted from the Parent-Child Conflict Tactics Scale (Straus
et al XXXXXXXXXXwere asked in relation to the selected child in the
past month. Types of punishments include spanking, hitting o
slapping on the bottom with a bare hand; hitting on the
ottom or elsewhere on the body with a hard object; hitting o
slapping on the face, head or ears; hitting or slapping on the
hand, arm or leg; and beating the child up. Along with
engaging in learning activities, making books and othe
learning material available for the child and avoiding harsh
disciplinary are the main responsibilities of the caree
(Bornstein et al. 2012).
Demographic characteristics of each household membe
were collected using structured questions. Among these, child
sex and age, living in a rural or an u
an area, the number of
children aged under five years in the household, maternal and
paternal education levels, and whether or not the mothe
and/or father were living at home were used in this study.
416 T. D. Tran et al.
© 2016 John Wiley & Sons Ltd, Child: care, health and development, 43, 3, 415–426
A nation’s Human Development Index (HDI) is a proxy
indicator developed by the United Nations Development
Program (UNDP) (UNDP XXXXXXXXXXEach country’s HDI was
obtained from the UNDP’s Human Development Reports
2011 (United Nations Development Programme (UNDP)
2011). The HDI ranges from 0 (the lowest) to 1 (the highest)
and is classified into very high (>0.790), high (>0.698 to
0.790), medium (>0.510 to XXXXXXXXXXand low categories XXXXXXXXXXo
less).
MICS procedures
The MICS data collection protocols are described in detail
elsewhere (United Nations Children’s Fund (Unicef)’s 2011;
Tran et al XXXXXXXXXXIn short, all MICS data used in this study were
collected through face-to-face structured interviews conducted
during home visits by national data collection teams. Data about
a child and caregiving practices were normally obtained from
the mother or primary caregiver of the child.
Statistical analysis
The estimations of means and percentages were calculated using
Stata’s survey commands in STATA Version 12 (Stata Corp
2011) taking into account cluster effects and sampling weights.
The hypothesized model of household wealth, caregiving
practices and child development (Fig. 1) was tested simulta-
neously using structural equation modelling in Mplus Version
7.3 (Muthén & Muthén XXXXXXXXXXIn the structural equation
model, the composite index of care for the child at home is
generated from the number of early learning activities engaged
in by the mother and the father, having 3+ children books at
home and having learning materials bought from shops. The
effect of family poverty on early childhood development via
unmeasured factors such as biological factors was treated as the
direct pathway from family poverty to the child outcome in
that model. Covariates were added into that model including
child sex and age, living in a rural or an u
an area, the
number of children aged under five years in the household,
maternal and paternal education levels and whether or not the
mother and/or father were living at home.
Structural equation modelling using multiple group analysis
with subgroups of HDI (low-HDI, medium-HDI and high-
HDI countries) was conducted to construct the same structural
equation model for each HDI group. The model coefficients
are interpreted as linear regression coefficients for the paths to
continuous outcomes. Model coefficients of the paths to binary
outcomes are odds ratios which were derived from original
probit regression coefficients for more straightforward inter-
pretation. Please see Supplementary File 1 for furthe
information about the structural equation modelling.
Results
Sample
Data on child development were collected in 44 countries at
MICS Round Four, but for nine countries data were not
available for public use by 1 April 2016. Overall, XXXXXXXXXXchildren
aged 36 to 59months from 35 countries in low, medium and
high-HDI groups were included in this study (Table 1). Of these,
the mean age was 47.1months (standard deviation, SD,
6