Talk:Ecological fallacy

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Philosophy (Rated Start-class, Mid-importance)
WikiProject icon This article is within the scope of WikiProject Philosophy, a collaborative effort to improve the coverage of content related to philosophy on Wikipedia. If you would like to support the project, please visit the project page, where you can get more details on how you can help, and where you can join the general discussion about philosophy content on Wikipedia.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.
WikiProject Statistics (Rated Start-class, Low-importance)
WikiProject icon

This article is within the scope of the WikiProject Statistics, a collaborative effort to improve the coverage of statistics on Wikipedia. If you would like to participate, please visit the project page or join the discussion.

Start-Class article Start  This article has been rated as Start-Class on the quality scale.
 Low  This article has been rated as Low-importance on the importance scale.

Origin of the name[edit]

Can someone point to the origin of this phrase? An attribution would be helpful, even if its "I made this up".

  • It's in zillions of books, so I don't think "I made this up" will be the answer. Michael Hardy 23:44, 11 Mar 2005 (UTC)
  • This article needs some references to back it up. 20 May 2005
  • C Achen & W P Shively (1995 book "Cross-level inference") and G King (1997 book "A Solution to the Ecological Inference Problem") provide histories that might be helpful. The term and its discussion start around the 1920s but renew as a discussion of analytic error in the 1950s. JeremyToday 01:38, 12 September 2005 (UTC)
  • If the term and its discussion start around 1920, why does the article say that it comes from a paper in 1950 (which was perhaps the "analytical error" version mentioned above)? If William Robinson really came up with the term (or even made it popular), he probably merits a page of his own, especially since it is in "zillions of books" (instead of a link to a disambig page which lists several people by that name who are NOT him), but if he is only one of the users of something earlier, I say we kill the link from his name. --Cromwellt | Talk 22:14, 28 October 2005 (UTC)

Simpson's paradox[edit]

How is the ecological fallacy different from Simpson's paradox? It seems like a link would at least be justified. Gray 21:07, 3 March 2006 (UTC)

Quite a fun description in example two. It's an example of 'ecological bias'. One thing I have noticed that is worrying on wikipedia is people tend to look at one phenomena and assume it is the same as another. It's like telling an academic that the term he is using means something else.

When you find two terms that appear to be alike the solution is not to merge the terms or relace one with another, the solution is to discover why two terms exist by conducting research! Whhat is being demonstrated is what I would called 'wikipedia subject bias' where the definition of terms on wikipedia is biased toward the subject discipline of wikipedia users with a complete disregard for the academic community as a whole from whom these 'rejected terms' originate. These terms exist for a reason! If one investigates why, one might discover a plethora of encyplopedic information....

Supposed 01:25, 10 May 2006 (UTC)

  • I believe that Simpson's paradox and the Ecological fallacy are identical. The "ecological fallacy" first appeared in the sociological literature in 1950, but Simpson's paradox was first described by Karl Pearson in 1899. Perhaps we need to merge the two articles. — Aetheling (talk) 20:20, 27 March 2008 (UTC)

TV Shows...[edit]

Guys the TV show examples (not that any happen in both shows) are distracting please change names and prune irrelevant information...


I don't think "constracted" is a word... but I'm not enough of a statistician to figure out what the right one is. (Contrasted with? Constricted to?) -- 21:20, 23 October 2006 (UTC)

Typical divide[edit]

The article said: "Within each community there is a typical divide between the rich and poor, the rich living in gated communities on the hills and the poor living adjacent to the industrial districts that pump carcinogens into their backyards." Is this a 'typical' divide? I don't think so. As a matter of fact, I think it is fairly unusual that rich people live in gated communities on hills. It might be typical in the country and time that the author live in, but I find it hard to believe that it is typical internationally and historically. I took the liberty of removing the word 'typical'. -Sensemaker


The section explaining the origin of the term is wrong: the problem described is not an ecological fallacy, but a reversal of cause and effect: literacy and immigration do correlate, it's just that a higher level of literacy attracts more immigration because of a more advanced economy, more jobs etc (instead of the reverse, that is, immigration bringing literacy, which is the conclusion the paper is attacking). This statistic can be propagated to individual level as well, showing that an individual immigrant is more likely to live in a high-literacy state. An ecological fallacy would be to find that immigrants actually live in low literacy states. Alfio 22:28, 18 March 2007 (UTC)

Reference 14 (the correction by Manfred Te Grotenhuis et al.) seems to have reversed signs (−0.53, −0.46 and +0.12) compared with the present Wikipedia article. — Preceding unsigned comment added by (talk) 13:32, 4 October 2012 (UTC)

The original paper matches the correction. The sign changes in the article were because illiteracy had been changed to literacy. To avoid similar confusion on the part of others, I've changed the article (without changing its meaning) to match the references. — Preceding unsigned comment added by (talk) 13:44, 4 October 2012 (UTC)

Article quality[edit]

I rate this article "Start" class at best, due to serious errors and ambiguities in the examples and definitions given for the ecological fallacy, and the lack of a figure or two to help readers understand how the fallacy operates. As time permits I will try to remedy the deficiencies, and I would welcome some help in this. — Aetheling (talk) 19:54, 27 March 2008 (UTC)

Statistical Issues[edit]

The ecological fallacy is not related to sampling per se. This point needs to be edited. Also the distinction between "biased sampling" and "stratified sampling" is vague and incomplete. A citation needs to be provided. Mattjans (talk) 15:54, 11 July 2008 (UTC)

EXACTLY. Here, we have too much math and too little reasoning and logic. It's a poor article in its current shape.  — Preceding unsigned comment added by (talk) 01:52, 20 April 2014 (UTC) 

Composition and Division fallacies[edit]

The "hasty generalization" fallacy is described accurately here as drawing a general conclusion based on too-little evidence. However, I don't think that's the converse of the environmental fallacy.

One of the college rhetoric texts I have used (I'll try to find it here shortly) defines the fallacies Division and Composition in much the same way as the Ecological fallacy (and its inverse). The fallacy of Division means to draw a conclusion about the individual members of a class based on attributes of the class. The Ecological fallacy example given here, the 2004 Washington State gubernatorial ballot dispute, seems to fit the definition of the Division fallacy well. Another example would be to conclude that any selected player on a winning team must be a winning player. The Ichiro batting average example also illustrates this fallacy.

The converse of Division, according to the text I was using, is Composition: drawing a conclusion about the whole (class) based on an attribute of its parts (members). The attribute may apply generally to all of the members of the class. This is different from Hasty Generalization, which is to jump to a conclusion based on too-little evidence: an insufficient subset of the class. Although these fallacies are similar, the key difference is that Composition might draw a conclusion about an attribute that applies to all members of the class, whereas Hasty Generalization does not consider the majority of the possible evidence (i.e., concluding that the class must have an attribute that one or a few members of the class have).

An example of the Composition fallacy would be to argue that a team made up of all the most-skilled players would be the best team. The New England Patriots' recent streak of Superbowl victories shows that this conclusion is a fallacy: the members of the team were not necessarily the ones who ran the most yards in the NFL that year, or had the most receptions - but as a team, the group performed very well together: an attribute of the class that is not necessarily an attribute of its members. The whole, in other words, can have attributes that are not found by a simple sum of the parts. (Likewise for the Division fallacy: the parts can have attributes that are not attributable to the whole.) Memetics (talk) 01:19, 29 July 2009 (UTC)

I searched Google Books for "ecological fallacy" "fallacy of division" to see if any reputable sources made a distinction between these two terms and discovered that most authors explicitly state they're equivalent errors. I went ahead and mentioned both fallacy of division and fallacy of composition (as the inverse), but we might want to merge this article with fallacy of division to avoid redundancy. —Mrwojo (talk) 02:21, 17 December 2010 (UTC)

Emile Durkeim's book on suicide.[edit]

In reference to the "citation needed" post made on Emile Durkeim's book on suicide: The wikipedia entry for Emile Durkeim contains some citations on this. Perhaps this is what the original poster was looking for. —Preceding unsigned comment added by Robinhoode (talkcontribs) 18:33, 7 November 2010 (UTC)

Ecological fallacy and Stereotypes[edit]

The article states that "Stereotypes, which assume that groups are homogeneous, are one form of ecological fallacy." But stereotypes need not be predicated on aggregate data, where across group differences can obscure estimates owing to group aggregation, which is an essential feature for the ecological fallacy. Stereotypes arise from the the blocking of data owing to prejudice and possibly a desire for simple ques to categorize reality.

In the example given: that a particular group of people have a lower IQ than the general population, the expected value of a random member of the particular group would have a point estimate lower than the general population. It may not be the case, owing to chance, but it is not due to error in the mean estimation owing to the effects of group aggregation . The stereotype essentially assumes a sampling error of zero, or perfect homogeneity in the population, in spite of variance in the data, and applies this assumption even when confronted with contradictory evidence.

At least so it appears to me, and I welcome correction. — Preceding unsigned comment added by Tachypaidia (talkcontribs) 02:30, 13 September 2011 (UTC)

Agree. For a binomial distribution, the mean and median are identical, aside from the trivial difference of having to take an integer value for the median, which is of no consequence in a large population. Thus, the probability that a randomly selected member of the population of interest will possess a characteristic that is known to exist in that population with some probability p, will indeed be p. This is no different than the classical example of randomly drawing a marble from each of two urns, the first containing a million marbles, of which 900,000 are green and 100,000 are red, the second 100,000 green and 900,000 red. Yes, you are more likely to get a green marble from the first urn than from the second, no doubt about that, and nobody but a fool, or possibly a student who has taken a "logical fallacies" class where the ecological fallacy has been watered down to a pablum suitable for dufuses, would bet any other way.

I certainly agree that we should judge people individually and not rashly, but the ecological fallacy does not provide an argument for doing so, and the example given is demonstrably bilgewater. (talk) 17:03, 16 March 2014 (UTC)

And while we are on this subject, we can observe that the example under "legal applications" is also flawed, since it deals with a binomial distribution and the statement by the "expert" in the case that "this approach was like trying to figure out Ichiro Suzuki's batting average by looking at the batting average of the entire Seattle Mariners team" is clearly inapplicable. Practically that case is about sampling bias. Nothing to do with the ecological fallacy. That some dufus judge thought it had something to do with the ecological fallacy does not mean that it did. TwoGunChuck (talk) 17:28, 16 March 2014 (UTC)


With all due respect to the editors, this article is nonsense, and it is the kind of nonsense that Wikipedia is often victim to. I do not mean that the content is nonsensical or even wrong. Rather, the idea that this should be an article is nonsense.

The article strings together some views about logical fallacies that occur in discussion of environmental issues. The logical fallacies themselves are already well covered in other Wikipedia articles. Singling our environmental issues as a hotbed of these fallacies is—a fallacy, and the kind of fallacy that others can exploit politically. These logical fallacies crop up everywhere, so it is misleading to suggest that environmental rhetoric is worse that other rhetoric when it comes to blunders in logic.

In a sense, this article is a double content fork, both from the many Wikipedia articles about logical fallacies and from the ecology articles to which these fallacies pertain. In my opinion, the appropriate solution is to edit existing Wikipedia articles on the pertinent ecological and environmental subjects, by adding internal links to existing Wikipedia articles on logical fallacies, and adding this article's reference citations—unless those ecology and environment articles already have these materials.

If an article like this belongs in Wikipedia, so do articles on Political fallacy, Economic fallacy, Romantic fallacy, Republican Party fallacy, Left Wing fallacy, Legal fallacy, Parenting fallacy, ad nauseum. Oh, and don't forget Regression analysis in quantitative economic theory, Programmable high-rise elevator controllers using fuzzy logic, Partial differential equations in fluid mechanics, Spellchecking poetry, Spellchecking blogs, Spellchecking newspaper articles, Spellchecking magazine articles, Spellchecking articles, Spellchecking technical manuals, Spellchecking refrigerator manuals, Spellchecking high school papers, Spellchecking junior high school papers, Spellchecking academic papers in anthropology, ..., ..., ... —Finell 02:57, 19 May 2014 (UTC)

Still, ecology has it's own distinct brand of fallacies, and you will find your work cut out trying to persuade the ecological community otherwise. Perhaps the article would be better renamed "Ecological fallacies". I like some of your ideas for new articles, particularly "Economic fallacies" and "Republican Party fallacies". --Epipelagic (talk) 03:54, 19 May 2014 (UTC)
Not according to the article itself. The article defines ecological fallacy, its subject, as a logical fallacy in interpreting statistical data. None of the so-called ecological fallacies is unique to ecology. Although the article mis-describes Simpson's paradox as one of the "four common statistical ecological fallacies", the reverse is actually true: fallacious conclusions in ecology based on Simpson's paradox are examples, or a subpopulation, of the fallacies that result from Simpson's paradox. The same is true of the other three "common statistical ... fallacies". Such fallacies abound across all subject matter because relatively few people understand logic, and far fewer have any clue about statistics; you can chalk the latter up to widespread innumeracy.
I'm glad you like the idea of "Republican Party fallacies". I hope you would appreciate equally a discussion of Democratic Party fallacies.—Finell 07:49, 24 May 2014 (UTC)
The argument seems to be that failures of reason common to a particular discipline do not deserve wikipedia pages to themselves. While I would broadly agree, this comment seems to have nothing to do with the article - unless all you knew about the article was its title. Finell wrote, "The article strings together some views about logical fallacies that occur in discussion of environmental issues," yet environmental issues appear nowhere in the article (nor were they in May 2014).
The term "ecological fallacy" is relevant to many disciplines, not just ecology. It is a general problem relevant to interpreting population level data; such datasets can be found across disciplines. Moreover, the term "ecological fallacy" is often found in technical literature, hence deserves an explanation in wikipedia. Flies 1 (talk) 18:31, 8 January 2016 (UTC)

mean, median, normal distribution[edit]

"Similarly, if a particular group of people is measured to have a lower average IQ than the general population, it is an error to conclude that a randomly selected member of the group is more likely to have a lower IQ than the average general population" It should be noted that if the group size becomes large enough it no longer is a fallacious conclusion - in the normal distribution, for large enough sample size, median equals mean and it is perfectly symmetrical around those. — Preceding unsigned comment added by (talk) 06:47, 1 October 2015 (UTC)

If you look at the example in the page with Group A and Group B, the inference that "a randomly selected member of the group is more likely to have a lower IQ than the average general population" is going to be wrong no matter how big the groups are. The central limit theorem would apply to measurements of the mean test scores for the two groups, or, loosely speaking, to aggregate properties of the groups; comparisons of randomly selected individuals in the respective groups will not approach a normal distribution no matter how big the groups are. As a matter of fact, the above unsigned comment seems to commit an ecological fallacy. Flies 1 (talk) 18:43, 8 January 2016 (UTC)