Ecological fallacy is a logical fallacy wherein someone makes an ecological inference, meaning an inference about individual behavior based on aggregate data, on false grounds.
The term originates from a research paper by William S. Robinson where he published his studies regarding ecological correlations, such as the correlation between literacy rate and the proportion of immigrants in each of the US states.
This fallacy, which is also known as “ecological inference fallacy”, belongs to the category of formal fallacies.
Ecological fallacy is an error in the interpretation of statistical data wherein inferences about the nature of specific individuals are made based on aggregate data for the group.
In simpler terms, this fallacy occurs when someone falsely assumes that an individual member of a group has the average characteristics of the larger group.
Its logical form goes as follows:
- Group 1 has characteristic X.
- Person Y is a member of group 1.
- Therefore, person Y has characteristic X.
An example of the ecological fallacy would be to assert that someone, who comes from a country with one of the highest crime rates, must be a criminal based on the crime statistics of their place of origin. This is considered fallacious since the evidence we have does not guarantee that the person is a criminal, even if it shows that they are more likely to be one.
Similarly, if one were to infer that John Doe, who is a male, must necessarily be taller than Kate Doe (female) simply because they know that the average height of males is considerably higher than females, they would fall prey to the ecological fallacy.
Furthermore, this fallacy may sometimes be confused with fallacy of division, however, the difference is that the latter is not seen as a statistical fallacy.
This concept originates from a 1950 paper by William S. Robinson in which he studied the correlation between literacy rate and the proportion of the population born outside of the United States.
His figures showed that the greater the proportion of immigrants in a given state, the higher its average literacy. However, when looking at the data on the level of individuals, the immigrants were on average less literate than native citizens.
He concluded that the positive correlation at the state level must have occurred because immigrants were more likely to settle in states where the native population had a higher literacy rate.