Ecological Correlations and the Behavior of Individuals.¶
Robinson, W. S. (1950). Ecological Correlations and the Behavior of Individuals. American Sociological Review, 15(3), 351-357.
Cited by¶
16 citations across 16 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Fallacy Of Misplaced Concreteness
- This is the same structural critique behind the ecological fallacy, mereological confusion, and the use-mention distinction, with misplaced concreteness as the general parent.
This sourceFoundational statement of the ecological fallacy — inferring individual-level relations from aggregate data — a structural sibling of misplaced concreteness.
- This is the same structural critique behind the ecological fallacy, mereological confusion, and the use-mention distinction, with misplaced concreteness as the general parent.
- Micro Macro Linkage
- And the paired fallacies: inferring micro from macro alone is the ecological fallacy, inferring macro from micro alone is the atomistic fallacy, and each is the structural-error diagnosis for ignoring one edge.
This sourceThe classic statement of the ecological fallacy (inferring individual-level from aggregate-level relations); its mirror is the atomistic fallacy.
- And the paired fallacies: inferring micro from macro alone is the ecological fallacy, inferring macro from micro alone is the atomistic fallacy, and each is the structural-error diagnosis for ignoring one edge.
- Modifiable Areal Unit Problem
- For two variables \(x, y\) over a base space of point-level observations, partitioned into cells, the total covariance decomposes as \(\text{Cov}(x, y) = \text{Cov}_{\text{between}}(\bar{x}, \bar{y}) + E[\text{Cov}_{\text{within}}(x, y)]\) — a between-cell term over the cell means plus a within-cell term averaged across cells.
This sourceShows that correlations computed on aggregated units differ systematically from individual-level correlations, the foundation of the within/between variance decomposition and the ecological fallacy.
- For two variables \(x, y\) over a base space of point-level observations, partitioned into cells, the total covariance decomposes as \(\text{Cov}(x, y) = \text{Cov}_{\text{between}}(\bar{x}, \bar{y}) + E[\text{Cov}_{\text{within}}(x, y)]\) — a between-cell term over the cell means plus a within-cell term averaged across cells.
- Neighborhood
- Global properties — connectedness, whole-system stability, fairness across an entire population — are not, and pretending they are by reasoning from a single neighborhood is a named error: the fallacy of composition, the ecological fallacy.
This sourceThe classic statement of the ecological fallacy — illegitimately promoting an aggregate/local-level conclusion to the individual level — the named error of gluing or escaping neighborhoods improperly.
- Global properties — connectedness, whole-system stability, fairness across an entire population — are not, and pretending they are by reasoning from a single neighborhood is a named error: the fallacy of composition, the ecological fallacy.
Domain-specific¶
- Atomistic Fallacy
- Ecological Correlation
- The critical fact, demonstrated formally by W. S. Robinson in 1950
This sourceUses Robinson's definition, in which an ecological correlation's statistical object is a group of persons - the unit of analysis, not the subject matter. Robinson's 1950 paper 'Ecological Correlations and the Behavior of Individuals', whose 48-state nativity/illiteracy coefficients are the worked ecological-correlation example. Robinson's cross-state ecological coefficient between nativity and illiteracy: -.526 on a state basis (-.619 by division), i.e. about +0.53 when re-signed for literacy.
SupportedVerified against a saved copy of the source
“certainly not equal to its corresponding individual correlation”
- The critical fact, demonstrated formally by W. S. Robinson in 1950
- Fallacy of Division
- The statistical form of this error has a landmark case: W. S. Robinson's 1950 paper "Ecological Correlations and the Behavior of Individuals."
This sourceThe landmark demonstration that correlations computed on aggregates need not match, and may reverse, the corresponding individual-level correlations.
SupportedVerified against the source
- The statistical form of this error has a landmark case: W. S. Robinson's 1950 paper "Ecological Correlations and the Behavior of Individuals."
Mechanisms¶
- Category-Membership Attribution Audit
- It is the concrete guard against the ecological fallacy, inferring an individual's traits from group aggregates
This sourceDistinguishes group-level ecological correlations from individual-level relationships and warns that the former cannot substitute for evidence about individuals.
- It is the concrete guard against the ecological fallacy, inferring an individual's traits from group aggregates
- Ecological Fallacy Guardrail
- That leap is exactly the ecological fallacy, and it can point the wrong way entirely: an aggregate correlation between two group properties can coexist with a null or opposite relationship at the person level (a district may have both more immigrants and more long-resident crime for unrelated reasons).
This sourceShows that ecological correlations between group-level rates need not correspond to, and may differ sharply from, correlations among individuals.
- That leap is exactly the ecological fallacy, and it can point the wrong way entirely: an aggregate correlation between two group properties can coexist with a null or opposite relationship at the person level (a district may have both more immigrants and more long-resident crime for unrelated reasons).
- Layered Risk Heatmap
- And a cell describes the unit, not each person in it — reading individual risk off an aggregate cell is the ecological fallacy.
This sourceDistinguishes group-level ecological measures from individual-level relationships and warns against inferring the latter from the former.
- And a cell describes the unit, not each person in it — reading individual risk off an aggregate cell is the ecological fallacy.
- Macro-to-Micro Operational Translation
- The naive move — tell every station to hit 85% — is precisely the error the method exists to prevent: a network average does not mean each unit faces the same condition.
This sourceDemonstrates that relationships measured for aggregate areas cannot safely be inferred for the individuals within them.
- The naive move — tell every station to hit 85% — is precisely the error the method exists to prevent: a network average does not mean each unit faces the same condition.
- Micro-Macro Crosswalk
- The classic misuse is running an inference the wrong way down it — reading the aggregate as if it described each member, the ecological fallacy
This sourceDemonstrates that aggregate correlations cannot generally substitute for individual-level relationships.
- The classic misuse is running an inference the wrong way down it — reading the aggregate as if it described each member, the ecological fallacy
- Micro / Meso / Macro Analysis
- Its classic failure is the ecological fallacy — reading a correlation that holds at the macro (aggregate) rung
This sourceShows why a correlation among aggregate groups cannot be read as the corresponding correlation among individuals.
- Its classic failure is the ecological fallacy — reading a correlation that holds at the macro (aggregate) rung
- Policy-Scale Analysis
- Its failure mode is the ecological fallacy
This sourceDefines the ecological inference error as using group-level averages or correlations to infer what is true of individuals.
- Its failure mode is the ecological fallacy
- Variability Analysis
- One is the ecological fallacy: reasoning from a group average down to individuals, or letting aggregation hide a reversal that appears within subgroups.
This sourceWarns that aggregate correlations do not support inference about individual behavior.
- One is the ecological fallacy: reasoning from a group average down to individuals, or letting aggregation hide a reversal that appears within subgroups.
- Vulnerability Index Construction
- And because the score attaches to an area or group, not a person, reading individual risk off it is the ecological fallacy.
This sourceAmerican Sociological Review. Shows that ecological variables describe groups rather than persons and that ecological correlations cannot be read as individual correlations.
- And because the score attaches to an area or group, not a person, reading individual risk off it is the ecological fallacy.
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Read against the text for 2 of 16 citations: 2 supported. Each verdict is shown under its citation below, with what in the work backs the sentence.
Was it audited? Yes. A second, independent pass read the citation against the article text and recorded a verdict.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Registry ID ref:eac8df6e14e6 · see in the full table