Skip to content

Geary's C

Measure global spatial autocorrelation by comparing weighted squared differences between neighboring observations with overall variance.

Version
v2 · 2026-09-19 · History
Domain-specific #
4675
Domain group
Formal Sciences
Origin domain
Experimental Design & Statistics
Subdomains
Spatial Statistics, Global Spatial Autocorrelation → Experimental Design & Statistics

Core Idea

Geary's C is a global spatial autocorrelation statistic formed from weighted squared differences across spatially linked observations, normalized by total variance. Neighbor differences contribute to the numerator, so similar adjacent values lower C and dissimilar adjacent values raise it relative to a random-spatial-association reference near one. The abstraction is therefore identified by a declared carrier, a transformation or constraint over that carrier, and an invariant that tells an analyst whether the named structure is genuinely present.

The load-bearing residual is not the broad topic of spatial statistics. It is the exact Geary normalization and neighbor-difference sensitivity, not spatial correlation generally.

How would you explain it like I'm…

Are Neighbors Alike?

Imagine a map of houses where each house has a number, like how many flowers it has. Geary's C is a score that checks if next-door houses have similar numbers. If neighbors are alike, the score is low; if neighbors are very different, the score is high; if it looks random, the score is about one.

The Neighbor Difference Score

Geary's C is a single number that tells you about a whole map at once. For each pair of places that count as neighbors, you look at how different their values are and square that difference. Then you compare the total to how spread out all the values are overall. If neighbors tend to be similar, Geary's C is below one; if neighbors tend to be different, it is above one; values scattered randomly give about one. You must say clearly which places count as neighbors, because that choice changes the answer.

Difference-Based Spatial Autocorrelation

Geary's C is a global measure of spatial autocorrelation: whether values at linked locations resemble each other across a whole map. It adds up weighted squared differences between neighboring values and divides by the overall variance, with weights saying which locations are linked and how strongly. Similar neighbors give small differences and push C below roughly one, the value expected under random spatial arrangement; dissimilar neighbors push it above one. Unlike Moran's I, which is built from products of deviations, Geary's C focuses directly on neighbor differences, so the two complement each other. A value below one is not by itself proof of correlation; you need a proper statistical test, and you must state the weights you used.

 

Geary's C is a global spatial autocorrelation statistic. Given values aligned to spatial units and a declared spatial weights matrix, its numerator is a weighted sum of squared pairwise differences between linked observations, and its denominator is the centered global sum of squares, with the standard normalization factors. Because it is difference-based, positive spatial association (similar neighbors) yields C below its null reference near one and negative association yields C above one. This neighbor-difference sensitivity is what distinguishes it from spatial correlation in general and complements Moran's I, which is cross-product based. Valid use requires publishing the weights and normalization, checking for zero variance and isolated units (islands) with no neighbors, and computing numerator and denominator consistently. Inference needs an explicit randomization (permutation) or asymptotic null; calling C < 1 'correlation' without that is overreach. Local Geary statistics are a different, location-specific tool and should not be substituted for the global coefficient.

Scope of Application

Geary's C belongs to spatial statistics and is useful where the analyst can specify values observed on spatial units plus a declared spatial weights matrix with zero diagonal, then evaluate C is computed from the declared weights and the ratio of weighted pairwise squared differences to centered global sum of squares. The scope is broad within that domain but bounded by the need for C is computed from the declared weights and the ratio of weighted pairwise squared differences to centered global sum of squares. The entry records a descriptive analytical identity; practical use requires the governing domain's evidence, standards, and safety obligations.

Clarity

The abstraction clarifies a crowded vocabulary by making C is computed from the declared weights and the ratio of weighted pairwise squared differences to centered global sum of squares the center of the account. A claim should name the carrier, the governing operation or relation, the applicable assumptions, and the recognition test. A bare label is insufficient because the name Geary's C can be used for a formal identity, an implementation, or a neighboring result unless carrier and convention are stated.

Manages Complexity

Without the abstraction, an analyst must reason directly over many local details: the carrier roles, admissibility assumptions, competing conventions, derived invariants, boundary cases, and proof or validation obligations specific to Geary's C. Geary's C compresses them into the roles in the structural signature. That compression permits comparison across instances without erasing the variables that determine validity. It also exposes which details may be varied safely and which are constitutive.

Abstract Reasoning

  1. Identify the carrier. State what the elements, states, objects, or observations are: values observed on spatial units plus a declared spatial weights matrix with zero diagonal. Reject examples whose alleged carrier belongs to a different problem.
  2. Lock the constitutive rule. Express C is computed from the declared weights and the ratio of weighted pairwise squared differences to centered global sum of squares independently of one notation or implementation.

Knowledge Transfer

Knowledge transfers strongly among subfields of spatial statistics because they reuse values observed on spatial units plus a declared spatial weights matrix with zero diagonal, Neighbor differences contribute to the numerator, so similar adjacent values lower C and dissimilar adjacent values raise it relative to a random-spatial-association reference near one., and publish the weights matrix and normalization, check zero variance and islands, compute numerator and denominator consistently, and use an explicit randomization or asymptotic null for inference. A theorem, diagnostic, or modeling warning can travel when those roles remain literal.

Relationships to Other Abstractions

Local relationship map for Geary's CParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Geary's CDOMAINPrime abstraction: Correlation — is a kind ofCorrelationPRIME

Current abstraction Geary's C Domain-specific

Parents (1) — outgoing typed relations

  • Geary's C is a kind of Correlation Prime

    The statistic literally measures association between values linked in space.

Hierarchy path (1) — routes to 1 parentless root

Neighborhood in Abstraction Space

Geary's C sits in a crowded region of the domain-specific corpus (39th percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.

Family — Spatial Statistics & Geographic Analysis (20 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-25