3-down/1-up Staircase

In a 3-down/1-up staircase, response history controls the next trial’s difficulty rather than treating every trial independently. Three successive correct responses trigger a harder stimulus, while a single incorrect response triggers an easier one. These opposing adjustments move presentation levels toward the region where the participant’s performance is stable enough to estimate perceptual sensitivity.

The three-correct versus one-incorrect rule sets the operating point of the staircase. Because difficulty increases only after a run of three correct responses but decreases after one error, the sequence converges near the intensity detected on approximately 79% of trials. Thus, the resulting estimate represents a defined performance level, not simply the participant’s average accuracy across all trials.

Accuracy is the key control signal, and stimulus intensity or task difficulty is the variable being adjusted. A participant’s correct and incorrect responses therefore determine whether subsequent trials become harder or easier. The changing sequence helps concentrate measurements near the relevant threshold, providing an efficient sensitivity estimate while limiting unnecessary testing and reducing participant fatigue.

Unlike a fixed-level design that presents selected stimulus intensities without adapting them to ongoing performance, the staircase uses each response to choose the next difficulty. This concentrates trials around the participant’s likely threshold instead of distributing testing across predetermined levels. That distinction explains why the method can reduce testing time while retaining a behavioral estimate of detection or discrimination ability.

A basic session presents a stimulus, records whether the participant responds correctly, and applies the rule to the next trial: three consecutive correct responses increase difficulty, whereas one incorrect response decreases it. Repeating this response-dependent sequence causes stimulus levels to approach the relevant performance region. The resulting pattern is then used to estimate the participant’s perceptual threshold.

In behavior research, the method can support detection, discrimination, and recognition measurements. Its use is not limited to one sensory modality: relevant applications include vision, hearing, touch, and other perceptual systems. Researchers can therefore apply the same adaptive logic to examine different forms of behavioral sensitivity when task performance is scored through correct and incorrect responses.