Written and maintained by CASRAI Editorial Board
Last updated
Short answer
Psychophysics is the set of methods for measuring how a physical stimulus relates to the sensation or response it produces. Gustav Fechner is credited as its founder. Its core quantities are thresholds: the smallest stimulus reliably detected and the smallest difference reliably noticed.
Psychophysics is the set of methods for measuring the relationship between a physical stimulus and the sensation or response it produces. A researcher presents a carefully controlled stimulus, asks a question whose answer can be turned into a number, and infers from those numbers how the sensory system behaves. Its core quantities are thresholds: the smallest stimulus a person reliably detects, and the smallest difference between two stimuli they reliably tell apart.
This page is written for people designing or reading a sensory experiment. Beyond the definitions, it sets the classical and modern methods side by side so you can see what each one estimates and where each is open to bias.
Definition and scope
Psychophysical methods are indirect: the experimenter cannot observe a sensation, so the method controls the stimulus, collects a quantifiable answer, and works backwards to what the mind did with the stimulus. Two basic questions organise the field. Detection asks how strong a stimulus must be before it is perceived. Discrimination asks how different two stimuli must be before the difference is perceived.
The approach is used across sensory neuroscience and experimental psychology, wherever the stimulus can be controlled and the response measured. It sits in the measurement toolkit alongside scaling approaches such as Thurstone scaling, and experiment-presentation software such as Inquisit is commonly used to run the trials.
Who founded psychophysics?
Gustav Theodor Fechner is credited as the founder. The classical psychophysical methods are described as developed as a result of Fechner’s book (Fechner, 1860), and one reference summary describes Fechner as working with Ernst Heinrich Weber in the nineteenth century. Weber’s work on how large a difference must be to be noticed is the precursor that Fechner built on, and it survives as Weber’s law.
What is the meaning of threshold in psychology?
A threshold is the boundary between a stimulus that is not detected and one that is. Two kinds matter most.
- Absolute threshold: the smallest stimulus, for example the smallest movement, that is reliably detected. A common reference definition is the least sensory input that can be detected, and by convention it is usually taken as the intensity detected at least 50 percent of the time.
- Difference threshold: the smallest difference between two stimuli that is detected reliably. This is the just noticeable difference (JND).
Because detection rises gradually with intensity rather than switching on at one value, a threshold is a statistical convention (such as the 50 percent point), not a fixed physical edge. That is why the way a study defines and estimates its threshold must be reported with the result.
Types or related concepts
The topic divides into three related families, each covered below: thresholds (absolute threshold and difference threshold, or JND), measurement methods (limits, adjustment, constant stimuli, staircase, forced choice, signal detection theory) and psychophysical laws (Weber, Fechner, Stevens). Detection and discrimination are the two basic questions that the threshold types answer.
Methods in practice
Classical methods
- Method of limits: the simplest method. Stimuli are presented in order from least to greatest intensity, or the reverse, with levels placed around the expected threshold.
- Method of constant stimuli: the same set of stimulus levels is presented in random order, and the proportion detected at each level is plotted, producing a curve that rises gradually.
- Method of adjustment: the participant changes the stimulus until it is just detectable; repeated settings are averaged.
- Staircase procedures: a refinement credited to Tom N. Cornsweet (1962) in which the stimulus moves up or down according to the participant’s responses, so fewer trials are needed.
Modern methods: forced choice and signal detection
Classical methods ask participants to report what they sense, which lets a participant bias the report, for example by being unwilling to say a faint stimulus was present. Forced-choice techniques were motivated by this problem. Signal detection theory goes further: it models the observer as comparing a noisy internal response against a criterion, saying “present” when the response falls above it. The distributions for noise alone and for signal plus noise overlap, so the criterion an observer adopts shapes the answer independently of the stimulus.
Method-selection table for experiment designers
Every cell below is limited to what the cited sources state.
| Method | What it does | Efficiency or bias note from the sources |
|---|---|---|
| Limits | Ordered ascending or descending stimuli around the expected threshold | Described as the simplest classical method |
| Adjustment | Participant sets the stimulus to just detectable; settings averaged | Relies on the participant’s own report of detectability |
| Constant stimuli | Fixed levels in random order; proportion detected per level | Yields a gradually rising detection curve, from which a 50 percent point can be read |
| Staircase | Stimulus adapts up or down to responses | Needs fewer trials |
| Forced choice | Participant must choose among alternatives | Introduced because reports in classical methods can be biased |
| Signal detection theory | Models noise, signal plus noise, and a response criterion | Separates the observer’s criterion from the stimulus effect |
How to read it: when the goal is a quick, rough threshold, the simpler classical methods are described as adequate; when a participant’s willingness to say “yes” could differ between conditions or groups, the forced-choice and signal-detection approaches exist precisely to stop that criterion from contaminating the threshold. Plan sample size for the comparison you intend to make with a tool such as G*Power, and state your hypothesis before collecting trials.
Psychophysical laws
- Weber’s law: just noticeable differences differ in size across stimulus levels.
- Fechner’s law: sensation grows with the logarithm of intensity; it works for brightness and loudness, but it is unclear that it holds for every sense.
- Stevens’ law: based on magnitude estimation; pain, for example, rises faster than Fechner’s law predicts.
Where it matters for labs, clinics or supply decisions
The practical consequences, drawn from the methods above, are about design and reporting rather than equipment. A reported threshold is only interpretable if the study states how it was defined (for example the 50 percent detection point) and which method produced it. Where a participant’s willingness to say “yes” could differ between conditions or groups, forced-choice or signal-detection designs keep that criterion from contaminating the result. Staircase procedures need fewer trials, which matters when session time is limited. Choice of presentation software or stimulus hardware should follow the method chosen, not the reverse.
Common misunderstandings
- “A threshold is a fixed number.” It is an estimate tied to a criterion such as 50 percent detection and to the method used.
- “Psychophysics measures perception directly.” It infers sensory processing from quantified responses.
- “Yes/no hit rates are pure sensitivity.” A participant’s criterion also changes the hit rate, which is why signal detection theory exists.
FAQ
Was the founder of psychophysics Fechner?
Yes. Gustav Fechner is credited as the founder, building on Ernst Weber’s earlier work; the classical methods followed his 1860 book.
What is the meaning of threshold in psychology?
The smallest stimulus (absolute threshold) or smallest difference between stimuli (difference threshold, or JND) that a person detects reliably, conventionally defined at about 50 percent detection.
What are the 7 steps of perception?
Psychophysics does not define a seven-step sequence. Stage lists of perception differ between textbooks, so use the list of the text your course or protocol follows. Psychophysics itself concerns measuring the stimulus-response relationship.
What are four types of perception?
Likewise, there is no single standard four-type list in psychophysics. Sources divide perception by sense modality or by theory, and the groupings vary; psychophysical methods apply within any single modality, such as vision, hearing or touch.
Sources
- Krantz, J. H., Experiencing Sensation and Perception, Chapter 2: Psychophysics, Hanover College: Chapter 2 (PDF) (read 2026-11-07).
- EBSCO Research Starters, Absolute Threshold (read 2026-11-07).
- Cornsweet, T. N. (1962). The staircase-method in psychophysics. American Journal of Psychology, 75(3), 485-491.
