AF Measure Analysis Issues I: Robustness analysis for the Alkire – Foster measures
|
|
Watch the video (includes video guide)
Key readings covered in this lecture
Atkinson, A. and Bourguignon, F. (1982), “The Comparison of Multidimensional Distributions of Economic Status”, Review of Economic Studies, XLIX, 183-201
Yalonetzky, G (2011), Conditions for the Most Robust Poverty Comparisons Using the Alkire-Foster Family of Measures. OPHI Working Paper 44b.
Alkire, S. Santon, M.E., Suman, S., and Yalonetzky, G. (2010). Is the Multidimensional Poverty Index robust to different weights? OPHI Working Paper 22a
Guide to the video
00:00 Introduction
02:20 Type of tools to analyse your measure – outline of the lecture
Part 1: Dominance Analysis (ordinal comparison)
05:01 Dominance conditions, the application on MPI
06:30 Three parameters affect ranking in the MPI (AF measures) – deprivation cut off, weights, poverty cut-off
07:47 Introduction of two possible evaluation methods of AF method
11:26 Stochastical dominance for the censored headcount and the adjusted headcount
12:10 Dominance conditions over k (Yalonetzky (2011)), proof, dominance results, new research
33:46 Results for dominance: union, intersection, intermediate.
Part 2: Rank Robustness
41:47 Other methods of rank robustness (Atkinson and Bourganoung (1982)), why we need other methods of robustness
44:43 Example of robustness of the global MPI, sensitivity to weights, correlation coefficients (Alkire et al (2010))
50:28 Multiple rank correlation
53:09 Explain rank correlation
54:11 Good and Kruskal’s correlation coefficient
56:45 Spearman’s correlation coefficient
59:26 Results of MPI rank robustness tests
62:57 Concordance indices for multiple ranking
69:10 Results of global MPI analysis
Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.
Calculating the M0 measure .do file
For supplementary reading on this topic view
