Open Conference Systems, 50th Scientific meeting of the Italian Statistical Society

Font Size: 
Modelling the effect of covariates for unbiased estimates in ecological inference methods
Venera Tomaselli, Antonio Forcina, Michela Gnaldi

Last modified: 2018-06-21


After showing that the estimates provided by three main ecological in- ference methods are heavily biased when compared to multilevel logistic models applied to a set of real individual data, the paper argues that ecological bias can be corrected only by accounting for relevant covariates. In addition, a data generating mechanism where bias cannot even be corrected by using covariates is described.


  1. Brown,P.J.&Payne,C.D.:Aggregatedata,ecologicalregression,andvotingtransitions.J. Am. Statist. Assoc. 81, 452–460 (1986)

  2. Cho, W. K. T.: If the assumption fits: A comment on the King ecological inference solution. Polit. Anal. 7, 143–163 (1998)

  3. Firebaugh, G.: A rule for inferring individual-level relationships from aggregate data. Am. Sociol. Rev. 43, 557–572 (1978)

  4. Forcina, A., Gnaldi, M., & Bracalente, B: A revised Brown and Payne model of voting be- haviour applied to the 2009 elections in Italy. Stat. Methods Appl. 21, 109–119 (2012)

  5. Freedman, D.A., Klein, S.P., Ostland, M., & Roberts, M.R.: On solutions to the ecological inference problem. J. Amer. Statist. Assoc. 93, 1518–1522 (1998)

  6. Gnaldi, M., Tomaselli, V., & Forcina, A.: Ecological Fallacy and Covariates: New In- sights based on Multilevel Modelling of Individual Data. I. Statist. Rev. 86, (2018) doi:10.1111/insr.12244

  7. Goldstein, H.: Multilevel Statistical Models. 4th ed. John Wiley & Sons, Chichester, UK (2011)

  8. Goodman, L. A.: Ecological regressions and behavior of individuals. Am. Sociol. Rev., 18, 351–367 (1953)

  9. Greiner, J. D. & Quinn, K. M.: R×C ecological inference: Bounds, correlations, flexibility and transparency of assumptions. J. Roy. Statist. Soc. Ser. A 172, 67–81 (2009)

  10. Hox, J. J., Moerbeek, M., & van de Schoot, R.: Multilevel Analysis: Techniques and Appli- cations. 2nd ed. Routledge, New York, NJ (2010)

  11. King,G.:ASolutiontotheEcologicalInferenceProblem:ReconstructingIndividualBehav- ior from Aggregate Data. Princeton University Press, Princeton NJ (1997)

  12. King, G.: The future of ecological inference research: A comment on Freedman et al.. J. Amer. Statist. Assoc. 94, 352–355 (1999)

  13. King, G., Rosen, O., & Tanner, M. A.: Binomial-Beta hierarchical models for ecological inference. Sociol. Methods Res. 28, 61–90 (1999)

  14. Liu,B.:EIextendedmodelandthefearofecologicalfallacy.Sociol.MethodsRes.20,1–23 (2007)

  15. Ng,K.W.,Tian,G.L.,&Tang,M.L.:DirichletandRelatedDistributions:Theory,Methods and Applications. John Wiley & Sons, Chichester UK (2011)

  16. Robinson, W. S.: Ecological correlations and the behavior of individuals. . Am. Sociol. Rev. 15, 351–357 (1950)

  17. Rosen, O., Jiang, W., King, G., & Tanner, M. A.: Bayesian and frequentist inference for ecological inference: The R×C case. Stat. Neerl. 55, 134–156 (2001)

  18. Snijders, T. A. B. & Bosker, R. J.: Multilevel Analysis: An Introduction to Basic and Ad- vanced Multilevel Modeling. Sage, London UK (2012)

  19. Subramanian,S.V.,Jones,K.,Kaddour,A.,&Krieger,N.:RevisitingRobinson:theperilsof individualistic and ecologic fallacy. Int. J. Epidem. 38, 342–360 (2009)

  20. Wakefield, J.: Ecological inference for 2×2 tables (with discussion).J. Roy. Statist. Soc. Ser. A 167, 1–42 (2004)

Full Text: PDF