Cue Validity¶
A cue's validity for a category is the chance of category membership among objects bearing that cue in a stated comparison population.
Core Idea¶
Cue validity for category c and feature f is P(c|f): the fraction of cue-bearing objects in a stated population that belong to c. It asks whether seeing the feature predicts category membership, not whether category members often have the feature. Rosch et al.'s original categorization work uses this direction; their category-level sum over cues is a score that can exceed one, not another probability.[^ref-327a13d012c5]
Scope of Application¶
Use it when features distinguish categories within a defined set of alternatives. Corter and Gluck's source-attested wings example shows the nested-category limit: P(animal|wings)≥P(bird|wings), so raw cue validity alone cannot mathematically select an intermediate basic level. In their explicitly hypothetical newspaper case, a 0.33 tabloid base rate, P(picture|tabloid)=0.9 and P(picture)=0.6 imply P(tabloid|picture)=0.495, while ubiquitous black ink leaves P(tabloid|black ink)=0.33. These cue-validity values are derived from their assumptions, not reported empirical observations.[^ref-a776575ca2cc]
Clarity¶
State sample, c, f and counts. In an explicitly invented population of 100 objects, 20 are chairs, 18 of those and 72 non-chairs have legs. Then P(chair|legs)=18/90=0.20, while P(legs|chair)=18/20=0.90. High within-category coverage is not high diagnosticity. P(c|f)−P© is a baseline-adjusted variant, not the defining conditional.[^ref-a776575ca2cc]
Manages Complexity¶
The number compresses feature-category co-occurrence and exposes how many other categories share a cue. It can hide correlations among cues, category nesting and sampling decisions. A summed category cue score must declare the feature set and must not be read as a probability.[^ref-327a13d012c5]
Abstract Reasoning¶
Form a two-by-two count of feature presence and category membership, divide joint c-and-f count by all f-bearing objects, then compare targets in the same population. Compute P(f|c) separately if feature coverage is the question. In cue selection, broad coverage can reduce discrimination when a feature is common outside c; compare both conditionals rather than collapsing them.[^ref-a776575ca2cc]
Knowledge Transfer¶
The conditional skeleton transfers to designed classifiers, but the cognitive category labels, feature coding and Rosch basic-level inference require new evidence. The original Rosch PDF was search-indexed but not text-readable in the web reader; no participant count or table value is invented here. The single-cue measure is strictly parented by Conditional Probability; the category-level summed score is not itself a probability.
[^ref-327a13d012c5]: Rosch et al., “Basic Objects in Natural Categories” (1976), original paper, scanned-access limit. [^ref-a776575ca2cc]: Corter and Gluck, “Explaining Basic Categories”, original analysis, pp.291–292.
Relationships to Other Abstractions¶
Current abstraction Cue Validity Domain-specific
Parents (1) — more general patterns this builds on
-
Cue Validity is a kind of Conditional Probability Prime
Single-cue validity is the conditional probability of a target category given an observed feature in a stated population.
Hierarchy paths (2) — routes to 2 parentless roots
- Cue Validity → Conditional Probability → Probability → Measure → Aggregation → Micro Macro Linkage
- Cue Validity → Conditional Probability → Probability → Measure → Set and Membership
Neighborhood in Abstraction Space¶
Cue Validity sits in a moderately populated region (58th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.
Family — Empirical Measurement & Statistical Inference Methods (50 abstractions)
Nearest neighbors
- Visual Capture — 0.86
- Join Count Statistic — 0.86
- Scientific Hypothesis — 0.85
- Occupancy–Abundance Relationship — 0.85
- Basic Category — 0.84
Computed from structural-signature embeddings · 2026-10-08