Tap to Read ➤

Explanation of Ecological Fallacy in Research with Examples

Neha B Deshpande Jan 23, 2019
Ecological fallacy is a fallacy in research, wherein you draw an inference about a group, and incorrectly attribute that inference to any individual in that group. This fallacy often leads to misleading conclusions. Here we explain its meaning with some examples.

Statistics Can Lie!

Numbers and figures are not always the correct representation of the truth. As a researcher, one has to carefully analyze data, and avoid falling prey to the ecological fallacy.
Research depends on a lot of statistical data. On the basis of this data, we form inferences and conclude the research. However, can statistics misguide an individual's research? Absolutely! That will happen if you assume that the observation regarding a group will hold true in case of a random individual picked from that group.

What is Ecological Fallacy?

Ecology is nothing but our 'environment'. Of course our environment influences us, but is it necessary that every individual tends to be similar to it? What if someone gives a prejudiced opinion about us on the basis of ecological references?
Ecological fallacy is a condition wherein you draw an inference about an individual on the basis of trends observed in the group to which he/she belongs. You tend to make a wrong assumption that, how the variables co-relate in an ecological level, they will do the same on an individual level.
Let us see some examples of Ecological fallacy.

Examples of Ecological Fallacy


In a class of 50 students, the average IQ level is high. However, if you pick a random student, that does not imply that he/she too will have a high level of IQ. Such an assumption is a case of ecological fallacy. It might happen that there are a few students who have an exceptionally high IQ.
Though many students have a very poor IQ level, you might end up with a misconception that just because a student belongs to that particular class, he has a good IQ level too.


Simpson's Paradox/Yule-Simpson Effect

This paradox implies that the inferences that hold true for different groups individually, hold completely untrue when it comes to the behavior of the two groups together. This difference in the behavior is known as the 'Simpson's paradox'. This paradox was presented by Edward H. Simpson in 1951.

E.g.(i) Passage of the Civil Rights Act of 1964.

Let us look at some examples of Simpson's Paradox.

Percentage of Democrats that voted in favor of the Act = 61%
Percentage of Republicans who voted in favor of the Act = 80%

A higher percentage of Republicans voted in favor of the Act. But when it was divided into the Northern and Southern States, the results were different.
Northern State
Percentage of Democrats that voted in favor of the Act = 94%
Percentage of Republicans who voted in favor of the Act = 85%

Higher percentage of Northern Democrats voted in favor of the Act.
Southern State
Percentage of Democrats that voted in favor of the Act = 7%
Percentage of Republicans who voted in favor of the Act = 0%

Higher percentage of Southern Democrats voted in favor of the Act.
Thus, both the statistics show different behavior when compared to a group and their respective individual performance, often misleading us. This is a classic real-life case of the Simpson's paradox.

E.g.(ii) Berkeley Gender Bias Case

The University of California, Berkeley, was accused of being gender biased, in 1973. According to statistics, the University had accepted 44% of male applicants and 35% of female applicants.

An analysis of the individual departments was made, which provided surprising results. These results proved to be mind-boggling, and antagonized the accusation made.
Data of six departments shows the following statistics.

Women Admitted:

Men Admitted:
An analysis proved that men had a tendency in seeking admission in those departments that had a higher rate of intakes. While women had a tendency to seek admission in departments with a lesser number of intakes. Thus the peculiarity of both genders and departments resulted in a misunderstanding that there was a gender bias in the admission process.
Thus, this shows a logical flaw. Exactly opposite of it is the exception fallacy, in which we tend to make a generalized statement about a whole group on the basis of a few exceptional individuals.