Summary Measure | Type | Usage in the analysis | Interpretation |
---|---|---|---|
Difference | Absolute-simple | Difference shows the absolute difference between two subgroups. It is used to analyse binary outcomes | A difference of zero shows no inequality. Summary measure greater than zero shows greater concentration of PFHIS coverage among the advantaged groups |
Ratio | Relative-Simple- | Ratio shows relative inequality between two subgroups. In this study, it is used to analyse binary outcomes | Ratio more than one show greater magnitude of inequality in comparison to the reference variable. Ratio is one in case of no inequality |
Absolute Concentration Index (ACI) | Absolute, Complex and ordered | Absolute concentration index is obtained by calculating the covariance between our health variable (PFHIS) and the individual’s rank in the social economic distributions. Further, this covariance is multiplied by 2 and divided by the mean level of health to get the summary measure | The ACI varies between -100 to 100 and positive values signifies concentration of PFHIS among the richer households. ACI of zero depicts no inequality |
Relative concentration index (RCI) | Relative, Complex and ordered | RCI, a complex and relative measure summarises inequality across social economic distributions that can be ordered. It is calculated by dividing the absolute concentration index by mean level of health and multiplied by 100 for better interpretation. The RCI is twice the area between the concentration curve (38) | The RCI varies between + 100 and -100 with zero depicting equal distribution and large positive values indicates higher coverage among the richer households and larger negative values indicating higher coverage among poor households |
Between Group Variance (BGV) | Weighted, complex, non-ordered dimension | Between group variance is calculated as the weighted sum of squared difference of the subgroup estimates and the setting average | BGV is always positive with higher values denoting higher inequalities. BGV |
Theil Index (TI) | Relative, complex, non-ordered dimension | Theil’s T statistic is population weighted and measures general disproportions. A subgroup’s population share (for eg. subgroup’s population / total weighted States population), the ratio of a state’s average PFHIS coverage among subgroup’s and overall average PFHIS coverage and the natural logarithm of this ratio, are multiplied; and then these products for each dimension are added to get a statistic. Estimates were multiplied by 1000 for better understanding | If there is no inequality, it takes the value zero. Greater absolute values indicate greater levels of inequality |