Ecological inference is a general statistical problem where a response variable is not available at the subject level because summary statistics are reported for groups only. It consists of merging information from different databases which are not linked to each other at the record level. We consider an election scenario where in each electoral precinct the fraction of voting-age people who turn out to vote, the fraction of black population and the number of voting-age people are observed. The proportions of blacks and of whites who vote are unobserved because electoral results and census data are not linked.
KeywordsAggregation Ecological inference Likelihood Markov chain Monte Carlo methods Method of bounds Nonparametric models Statistical approaches
- Achen, C., and W. Shively. 1995. Cross-level inference. Chicago: University of Chicago Press.Google Scholar
- King, G. 1997. A solution to the ecological inference problem: Reconstructing individual behavior from aggregate data. Princeton: Princeton University Press.Google Scholar
- King, G., O. Rosen, and M. Tanner, eds. 2004. Ecological inference: New methodological strategies. Cambridge: Cambridge University Press.Google Scholar
- King, G., O. Rosen, M. Tanner, and A. Wagner. 2003. The ordinary election of Adolf Hitler: A modern voting behavior approach. Online. Available at https://gking.harvard.edu/files/gking/files/nazivp.pdf Accessed 16 Aug 2006.
- Staniswalis, J. 2005. On fitting generalized non-linear models with varying coefficients. Computational Statistics and Data Analysis 50: 893–902.Google Scholar
- Wakefield, J. 2004. Prior and likelihood choices in the analysis of ecological data. In Ecological inference: New methodological strategies, ed. G. King, O. Rosen, and M. Tanner. Cambridge: Cambridge University Press.Google Scholar