Ecological Bias, Confounding, and Effect Modification Export

@article{greenland_and_morgenstern_1989, abstract = {Greenland S (Division of Epidemiology, UCLA School of Public Health, Los Angeles, California 90024, USA) and Morgenstern H. Ecological bias, confounding, and effect modification. International Journal of Epidemiology 1989, 18: 269-274. Ecological bias is sometimes attributed to confounding by the group variable (ie the variable used to define the ecological groups), or to risk factors associated with the group variable. We show that the group variable need not be a confounder (in the strict epidemiological sense) for ecological bias to occur: effect modification can lead to profound ecological bias, whether or not the group variable or the effect modifier are independent risk factors. Furthermore, an extraneous risk factor need not be associated with the study variable at the individual level in order to produce ecological bias. Thus the conditions for the production of ecological bias by a covariate are much broader than the conditions for the production of individual-level confounding by a covariate. We also show that standardization or ecological control of variables responsible for ecological bias are generally insufficient to remove such bias. 10.1093/ije/18.1.269}, author = {Greenland, Sander and Morgenstern, Hal}, citeulike-article-id = {3398882}, citeulike-linkout-0 = {http://dx.doi.org/10.1093/ije/18.1.269}, citeulike-linkout-1 = {http://ije.oxfordjournals.org/cgi/content/abstract/18/1/269}, citeulike-linkout-2 = {http://view.ncbi.nlm.nih.gov/pubmed/2656561}, citeulike-linkout-3 = {http://www.hubmed.org/display.cgi?uids=2656561}, comment = {Argues that ecological bias is a much bigger problem than confounding in individual level studies. It may be due to effect modification as well. The trouble is that it is impossible to establish from the data alone which of these it is. This, however, is only applicable to the practice of attempting to make individual-level inferences from ecological data.}, day = {1}, doi = {10.1093/ije/18.1.269}, journal = {Int. J. Epidemiol.}, keywords = {confounding, contextual, ecological-bias, effects}, month = {March}, number = {1}, pages = {269--274}, posted-at = {2008-10-13 02:47:45}, priority = {0}, title = {Ecological Bias, Confounding, and Effect Modification}, url = {http://dx.doi.org/10.1093/ije/18.1.269}, volume = {18}, year = {1989} }

TY - JOUR ID - greenland_and_morgenstern_1989 L3 - citeulike-article-id:3398882 N2 - Greenland S (Division of Epidemiology, UCLA School of Public Health, Los Angeles, California 90024, USA) and Morgenstern H. Ecological bias, confounding, and effect modification. International Journal of Epidemiology 1989, 18: 269-274. Ecological bias is sometimes attributed to confounding by the group variable (ie the variable used to define the ecological groups), or to risk factors associated with the group variable. We show that the group variable need not be a confounder (in the strict epidemiological sense) for ecological bias to occur: effect modification can lead to profound ecological bias, whether or not the group variable or the effect modifier are independent risk factors. Furthermore, an extraneous risk factor need not be associated with the study variable at the individual level in order to produce ecological bias. Thus the conditions for the production of ecological bias by a covariate are much broader than the conditions for the production of individual-level confounding by a covariate. We also show that standardization or ecological control of variables responsible for ecological bias are generally insufficient to remove such bias. 10.1093/ije/18.1.269 IS - 1 JF - Int. J. Epidemiol. EP - 274 TI - Ecological Bias, Confounding, and Effect Modification VL - 18 SP - 269 KW - confounding KW - contextual KW - ecological-bias KW - effects N1 - Argues that ecological bias is a much bigger problem than confounding in individual level studies. It may be due to effect modification as well. The trouble is that it is impossible to establish from the data alone which of these it is. This, however, is only applicable to the practice of attempting to make individual-level inferences from ecological data. AU - Greenland, Sander AU - Morgenstern, Hal PY - 1989/03/01/ UR - http://dx.doi.org/10.1093/ije/18.1.269 ER -