Article ID: | iaor20084552 |
Country: | Netherlands |
Volume: | 176 |
Issue: | 3 |
Start Page Number: | 1811 |
End Page Number: | 1838 |
Publication Date: | Feb 2007 |
Journal: | European Journal of Operational Research |
Authors: | Zanakis Stelios H., Alvarez Cecilia, Li Vivian |
Keywords: | statistics: data envelopment analysis, statistics: regression |
This paper examines the influence and direction of social and economic determinants of the HIV/AIDS global epidemic across nations and assesses each country's efficiency in battling the pandemic. The initial dataset consisted of 151 countries with five dependent variables and 90 explanatory variables (reduced to 50 after extensive exploratory data analysis of missing value patterns, undesirable multi-colinearities and multivariate outliers). Five measures were analyzed, namely HIV/AIDS Cases per 100,000 population; number and percentage of adults age 15–49 living with HIV/AIDS virus; estimated number of AIDS-related deaths for adults and children; and percentage of male sexual transmitted disease patients diagnosed with HIV/AIDS (identified by canonical correlation as not amenable to significant predictions). Reasonably good fit regression models (