Atomistic Fallacy

The inferential error of reading a within-individual relationship directly onto a group or population, ignoring the contextual variance operating only at the group level — the mirror image of the ecological fallacy.

Core Idea

The atomistic fallacy is the error of drawing conclusions about group- or population-level relationships solely from individual-level data. It is the mirror image of the ecological fallacy, running the projection in the opposite direction. The mechanism is level-nonpreservation: a within-individual correlation need not equal the cross-group correlation, because group-level outcomes are shaped by contextual forces — inequality, institutional environment, peer composition — that have no individual-level counterpart, and reading one onto the other is unsupported.

Scope of Application

The atomistic fallacy lives across the multilevel-inference subfields of statistics and the empirical disciplines working with hierarchical data, wherever a within-level estimate is read onto a different level carrying contextual variance.

• Social epidemiology — the home turf and source of the modern name (Diez-Roux 1998).
• Political science — the individual wealth-vote gradient reversing at the state level (Gelman's red-state paradox).
• Public-health intervention design — individual dose-response read onto populations carrying herd effects.
• Educational research — a student-level teaching effect need not survive to district policy.
• Multilevel modelling methodology — the standing rationale for multilevel models, taught paired with its dual.

Clarity

Naming the atomistic fallacy makes the symmetry of cross-level inference errors legible. The ecological fallacy is guarded against routinely; its mirror slips by under the false intuition that individual data is the "safer" ground. Pairing the two as duals dissolves that asymmetry of vigilance. The sharper question is "which level does my question actually live at?" — which dictates the design and localizes the gap to contextual variance a better within-level estimate cannot fix.

Manages Complexity

Cross-level inference is otherwise a case-by-case minefield, each substrate posing its own puzzle about whether a relationship survives aggregation. The fallacy compresses the space onto three things: the level observed, the level the question lives at, and whether contextual variance sits between them. Where they align, the estimate transfers; where levels differ with context present, expect divergence in magnitude, sign, or mechanism. Pairing with the dual collapses two vigilances into one symmetric rule.

Abstract Reasoning

The concept licenses diagnostic reasoning — detecting the error from a level mismatch plus contextual variance and predicting the form of failure. It supports interventionist reasoning that models both variance components rather than sharpening the within-level estimate, boundary-drawing that treats the fallacy and its ecological dual symmetrically and fixes which level the question lives at, and predictive reasoning that forecasts a strong within-level gradient weakening, vanishing, or reversing at the higher level.

Knowledge Transfer

Within multilevel statistical inference the fallacy transfers literally as mechanism across every field with hierarchical data — epidemiology, political science, education, organizational research — carrying its diagnostic, failure catalogue, and multilevel corrective, and travelling paired with its ecological dual. But this is one substrate. Beyond it the named diagnostic does not travel; the general "parts need not behave like wholes" structure recurs via aggregation, emergence, and the level-of-analysis machinery of statistical_inference. Non-statistical uses are analogy.

Relationships to Other Abstractions

Current abstraction Atomistic Fallacy Domain-specific

Parents (3) — more general patterns this builds on

• Atomistic Fallacy is a kind of Fallacy of Composition Domain-specific

  Atomistic Fallacy is Fallacy of Composition specialized to projecting an individual-level statistical relationship onto a group or population.

• Atomistic Fallacy presupposes Aggregation Prime

  Atomistic Fallacy presupposes a level-forming aggregation from individual observations to a group or population target.

• Atomistic Fallacy presupposes Statistical Inference Prime

  Atomistic Fallacy presupposes an estimate or relationship inferred at one statistical level and projected to another.

Hierarchy paths (10) — routes to 8 parentless roots

• Atomistic Fallacy → Fallacy of Composition → Cross-Level Inference → Hierarchy → Network → Reservoir-Flux Network → Conservation Laws → Invariance

Show alternative paths (9)

Neighborhood in Abstraction Space

Atomistic Fallacy sits in a sparse region of the domain-specific corpus (65^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Social Perception & Self-Referential Bias (23 abstractions)

Nearest neighbors

• Ecological Inference Problem — 0.85
• Ecological Correlation — 0.84
• Key Informant — 0.84
• Illusory Superiority — 0.82
• Kuznets curve — 0.82

Computed from structural-signature embeddings · 2026-07-12