Fig. 5

Gini Coefficient compared to Coefficient of Variation for inter-population comparisons. This shows strong correlation between the Gini coefficient and the standard coefficient of variation (standard deviation divided by the mean) illustrating the Gini coefficient as a robust general statistical descriptor of variation. This is alongside its interpretive utility as being bounded between 0 and 1 (whereas the coefficient of variation is an unbounded proportional value that does not offer an intuitive scale for epidemiological comparisons). There were 21 values compared for each index here, representing data from different analyses presented in this paper: eight for various diseases among world health regions (from Fig. 1), one for malaria across the countries of West Africa (Fig. 2), four of sub-national variation in malaria prevalence in selected West African countries (Fig. 3), and eight sequential surveys across villages in The Garki Project (Fig. 4)