Ecologic versus individual-level sources of bias in ecologic estimates of contextual health effects

被引:210
作者
Greenland, S [1 ]
机构
[1] Univ Calif Los Angeles, Sch Publ Hlth, Dept Epidemiol, Topanga, CA 90290 USA
[2] Univ Calif Los Angeles, Coll Letters & Sci, Dept Stat, Topanga, CA 90290 USA
关键词
aggregate studies; confounding; contextual studies; ecologic fallacy; ecologic studies; environmental health; epidemiology; multilevel studies; relative risk; risk assessment;
D O I
10.1093/ije/30.6.1343
中图分类号
R1 [预防医学、卫生学];
学科分类号
1004 ; 120402 ;
摘要
A number of authors have attempted to defend ecologic (aggregate) studies by claiming that the goal of those Studies is estimation of ecologic (contextual or group-level) effects rather than individual-level effects. Critics of these attempts point out that ecologic effect estimates are inevitably used as estimates of individual effects, despite disclaimers. A. more subtle problem is that ecologic variation in the distribution of individual effects can bias ecologic estimates of contextual effects. The conditions leading to this bias are plausible and perhaps even common in studies of ecosocial factors and health outcomes because social context is not randomized across typical analysis units (administrative regions). By definition, ecologic data contain only marginal observations on the joint distribution of individually defined confounders and outcomes, and so identify neither contextual nor individual-level effects. While ecologic studies can still be useful given appropriate caveats, their problems are better addressed by multilevel study designs, which obtain and use individual as well as group-level data. Nonetheless, such studies often share certain special: problems with ecologic studies, including problems due to inappropriate aggregation and problems due to temporal changes in covariate distributions.
引用
收藏
页码:1343 / 1350
页数:8
相关论文
共 72 条
[21]  
Freedman D. A., 1998, J AM STAT ASSOC, V93, P1518
[22]   ECOLOGICAL REGRESSION AND VOTING-RIGHTS [J].
FREEDMAN, DA ;
KLEIN, SP ;
SACKS, J ;
SMYTH, CA ;
EVERETT, CG .
EVALUATION REVIEW, 1991, 15 (06) :673-711
[23]   The future of ecological inference research: A comment on Freedman et al. - Response to King's comment [J].
Freedman, DA ;
Ostland, M ;
Roberts, MR ;
Klein, SP .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1999, 94 (445) :355-357
[24]  
Goldstein H., 2010, Multilevel statistical models, V4th
[25]   SOME ALTERNATIVES TO ECOLOGICAL CORRELATION [J].
GOODMAN, LA .
AMERICAN JOURNAL OF SOCIOLOGY, 1959, 64 (06) :610-625
[26]   When should epidemiologic regressions use random coefficients? [J].
Greenland, S .
BIOMETRICS, 2000, 56 (03) :915-921
[27]  
Greenland S, 1999, STAT SCI, V14, P29
[28]   ACCEPTING THE LIMITS OF ECOLOGIC STUDIES - REPLY [J].
GREENLAND, S ;
ROBINS, J .
AMERICAN JOURNAL OF EPIDEMIOLOGY, 1994, 139 (08) :769-771
[29]   NEITHER WITHIN-REGION NOR CROSS-REGIONAL INDEPENDENCE OF EXPOSURE AND COVARIATES PREVENTS ECOLOGICAL BIAS [J].
GREENLAND, S ;
MORGENSTERN, H .
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY, 1991, 20 (03) :816-817
[30]   ECOLOGICAL BIAS, CONFOUNDING, AND EFFECT MODIFICATION [J].
GREENLAND, S ;
MORGENSTERN, H .
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY, 1989, 18 (01) :269-274