The ecological fallacyBMJ 2011; 343 doi: https://doi.org/10.1136/bmj.d4670 (Published 27 July 2011) Cite this as: BMJ 2011;343:d4670
- Philip Sedgwick, senior lecturer in medical statistics
- 1Section of Medical and Healthcare Education, St George’s, University of London, Tooting, London, UK
Researchers explored the relation between morbidity from injury in children and socioeconomic deprivation. An ecological cross sectional study of 862 electoral wards in the Trent region of northern England was performed.1 For each electoral ward the researchers obtained rates of admission of children to hospital for all injuries between 1992 and 1997. All hospitals in the Trent region were included, and data were aggregated for each electoral ward, regardless of which hospital children were admitted to. The Townsend score associated with each electoral ward was obtained. This score assesses socioeconomic deprivation in families and includes measurement of employment status, overcrowding, car ownership, and owner occupation status.
A positive association between rates of hospital admission for all injuries in children and socioeconomic deprivation was reported, with those electoral wards with the greatest socioeconomic deprivation having the highest hospital admission rates. The researchers commented that the results were subject to the ecological fallacy.
Which one of the following statements best describes the ecological fallacy in relation to the results of this study?
a) Children with an injury may be admitted to a hospital in an electoral ward in which they do not live.
b) Only those hospitals in electoral wards with the greatest socioeconomic deprivation admit children with an injury.
c) Children from families with greater socioeconomic deprivation are more likely to be admitted to hospital with injuries.
d) Hospitals are located only in electoral wards with the greatest socioeconomic deprivation.
Answer c best describes the ecological fallacy regarding the results of the above study.
In this study the electoral ward was the unit of analysis. Data were collected for groups of people aggregated across the electoral ward and not for individuals within the Trent region. The ecological fallacy is a term used when data collected at a group level are analysed and the results assumed to apply to relationships at the individual level. The study reported a positive association between rates of hospital admissions for injury in children in electoral wards and socioeconomic deprivation associated with the ward. To assume that these results could be applied to the individual—that is, that children from families with greater socioeconomic deprivation are more likely to be admitted to hospital with injuries—is an erroneous belief termed the ecological fallacy (answer c). Although electoral wards with the highest rates of hospital admission for injury have the greatest socioeconomic deprivation, we cannot assume that children from families with the greatest socioeconomic deprivation are those children being admitted to hospital.
For each electoral ward the rate of hospital admissions for all injuries in children between 1992 and 1997 together with the associated Townsend scores were collated. The data collected were at the group level—that is, for electoral wards with no reference to where hospitals were located (answers a, b, and d are false).
The purpose of an ecological study is to make large scale comparisons between groups of people—for example, the health status of countries. Ecological studies do not allow examination of individuals and are therefore open to bias as described. However, they do allow an initial examination of the health status and needs of communities.
Cite this as: BMJ 2011;343:d4670
Competing interests: None declared.