CHAPTER THREE
METHODOLOGY
3.0 Introduction
This chapter outlines the methodological approach employed to examine the geospatial
distribution and determinants of child malnutrition in Nigeria. It details the research design,
study area, data sources, data collection techniques, and the analytical methods used.
Specifically, it describes the use of geospatial analysis and logistic regression to explore the
spatial variation and key predictors of malnutrition among children under five years old. Stata
was utilized for statistical and geospatial analysis.
3.1 Research Design
This study adopts a quantitative research design that integrates geospatial and statistical
techniques to analyze child malnutrition patterns in Nigeria. The study leverages secondary data
from the Nigeria Demographic and Health Survey (NDHS) 2018 to examine spatial disparities in
malnutrition prevalence and identify associated socio-demographic and environmental factors.
The analysis employs logistic regression to model the relationship between malnutrition and key
predictor variables, while geospatial techniques, including spatial interpolation and hotspot
analysis, are used to visualize the spatial distribution of malnutrition rates across Nigeria.
3.2 Study Area
Nigeria, the most populous country in Africa, serves as the study area for this research. The
country comprises 36 states and the Federal Capital Territory (FCT), Abuja. It spans
approximately 923,769 square kilometers, encompassing diverse geographical zones, from the
Sahel in the north to coastal rainforests in the south. With a population exceeding 200 million,
Nigeria's child malnutrition rates vary significantly across regions due to socio-economic,
cultural, and environmental factors.
Malnutrition, particularly stunting, wasting, and underweight, remains a major public health
issue in Nigeria, with variations linked to differences in food security, maternal health practices,
education, and access to healthcare.
3.3 Source of Data
The study relies on secondary data obtained from the 2018 Nigeria Demographic and Health
Survey (NDHS), conducted by ICF Macro in collaboration with the National Population
Commission (NPC) of Nigeria. The NDHS employs a multi-stage stratified cluster sampling
technique to ensure national representation.
The dataset includes anthropometric measurements such as height-for-age (stunting), weight-for-
height (wasting), and weight-for-age (underweight) for children under five years old.
Additionally, household and maternal characteristics, including maternal education, household
wealth index, access to healthcare, and geographic coordinates of survey clusters, are included in
the analysis.
3.4 Method of Data Analysis
3.4.1 Geospatial Analysis Geospatial techniques were used to analyze and visualize the spatial
distribution of child malnutrition across Nigeria. The geospatial analysis included:
Spatial Interpolation: Kriging and Inverse Distance Weighting (IDW) methods were
used to estimate malnutrition prevalence in unsampled areas based on survey data.
Hotspot Analysis: Getis-Ord Gi statistic was applied to identify statistically significant
clusters of high and low malnutrition prevalence.
Spatial Autocorrelation: Moran’s I statistic was used to measure the degree of
clustering in child malnutrition rates across different regions.
3.4.2 Logistic Regression Analysis
To identify the key socio-demographic and environmental factors influencing child malnutrition,
a binary logistic regression model was employed. The dependent variable was malnutrition
status, categorized as malnourished (1) or not malnourished (0), based on WHO-recommended
Z-scores for stunting, wasting, and underweight.
The logistic regression model is given as:
Logit ( p )=log ( 1−pp )=β + β x + β x +⋯+ β x
0 1 1 2 2 n n (3.1)
Where, р is the probability of the event occurring (dependent variable = 1), β 0 is the
intercept, and β 0 , β 1 ,⋯ , β n are the coefficients for the independent variables X 1, X 2 ,…, X n.
3.4.3 Spatial-Statistical Integration To enhance the interpretation of malnutrition determinants,
the results from the logistic regression were integrated into the geospatial analysis. The predicted
probabilities of malnutrition were mapped to visualize how socio-economic and environmental
factors contribute to regional disparities in child malnutrition.