Glossary
Weber-Fechner Law
What is the Weber-Fechner law?
The Weber-Fechner law concerns relationships between physical stimulus intensity and subjective sensation. It combines Weber's finding about detectable differences with Fechner's additional account of how sensation grows. These are related but distinct claims.
Just noticeable difference
A just noticeable difference (JND) is a difference threshold measured under specified conditions and a chosen performance criterion. It is not the smallest change a person invariably detects.
Weber's law
Weber's law proposes an approximately constant ratio between a detectable change and the starting intensity: ΔI/I = k. A larger starting stimulus therefore requires a larger absolute change for the same proportional difference, within the range where the model fits.
Here, I is the starting intensity, ΔI is the detectable change, and k is the Weber fraction. If k were 0.05 in a particular test, a starting intensity of 100 units would imply a 5-unit difference threshold; 200 units would imply 10 units.
Fechner's law
Fechner proposes that subjective intensity grows logarithmically with physical intensity. Deriving this from Weber's relation assumes that successive JNDs represent equal increments of sensation. That additional assumption does not hold in every setting.
Examples and applications
- Lighting and interface design: Detecting a brightness change depends on the starting level, adaptation, contrast, and viewing conditions. Test whether the intended difference is visible.
- Pricing: A fixed price change may seem smaller relative to an expensive product. This is an analogy; a sensory law does not establish a purchasing rule or a threshold for price changes.
Limits and Stevens's power law
Model fit depends on the sensory dimension, range, and measurement method. Stevens's power law is an alternative for subjective magnitude judgments.
Why it matters
The size of a physical change and the size of a perceived change are different questions. Psychophysics studies their relationship instead of assuming one is a simple copy of the other.