Decomposing Reaction Time - Which Stage Does Prediction Shorten?
Reaction time decomposes into perception (stimulus detection and recognition), cognition (response selection and decision), and motor (muscle contraction and movement). Simple reaction time (one stimulus type, one response type) comprises perception 30-50ms, cognition 50-80ms, motor 30-50ms, totaling 150-200ms. Prediction primarily shortens the cognitive stage. When stimulus timing or type is predictable, response selection processing pre-completes before stimulus onset, requiring only motor command initiation after detection. This compresses the cognitive stage from 50-80ms to 10-20ms, shortening total reaction time by 30-50ms. Express saccades (80-120ms) and trigger responses (100-130ms) result from maximally functioning predictive processing. However, when predictions are wrong, response inhibition and re-selection are needed, conversely adding 50-100ms delay.
Temporal Prediction and Preparatory Potentials
Predicting when stimuli will appear (temporal prediction) optimizes motor system readiness. EEG studies show Contingent Negative Variation (CNV), a slow potential over frontal-central areas, appearing just before predicted stimulus onset. Larger CNV amplitude correlates with shorter reaction time, reflecting motor cortex neurons in near-threshold ready state. Temporal prediction accuracy depends on stimulus interval regularity. With constant intervals, the cerebellum learns timing and accurately predicts next stimulus onset. Bench reaction time tests include interval randomness preventing perfect temporal prediction, but learning the statistical distribution of intervals enables partial prediction. Temporal prediction accuracy improves with trials, contributing to warm-up effects.
Spatial Prediction and Attention Pre-Allocation
Predicting where stimuli will appear (spatial prediction) shortens the perceptual stage by pre-allocating attention. In Posner's cueing paradigm, responses to stimuli appearing after valid cues (indicating correct location) are 30-50ms faster than after invalid cues. This occurs because visual processing is enhanced at pre-attended locations. Attention pre-allocation raises baseline activity of neurons in visual areas like V4 (color processing) and MT (motion processing), shortening response latency and increasing amplitude. Since Bench tests often have fixed stimulus locations, pre-allocating attention to the stimulus area before test start enables optimal response from the first trial. A 'broad attention' state gazing at screen center while distributing attention to peripheral vision is most efficient for unpredictable stimulus positions.
Probabilistic Prediction and Stimulus-Response Association Learning
In choice reaction time tasks, when stimulus probabilities are unequal, responses to high-frequency stimuli are 20-40ms faster than low-frequency ones. This probabilistic prediction effect occurs because the brain implicitly learns stimulus probabilities and pre-prepares high-probability responses. This learning is handled by the striatum in basal ganglia, adjusting stimulus-response association strength through reward prediction error signals. In Bench tests, when specific patterns or regularities exist (e.g., color change patterns, character frequency), repeated trials implicitly learn these statistical regularities, accelerating responses. This learning proceeds automatically without conscious effort, but directing attention to test structure can accelerate learning speed. However, excessive pattern searching consumes attention resources and can backfire.
Practical Approaches to Improve Prediction Accuracy
Practical approaches for improving prediction ability and shortening reaction time. First, promoting implicit learning through test repetition. Repeating the same test automatically learns stimulus timing distribution, spatial patterns, and probability structure. Learning progresses rapidly in the first 10-20 attempts, then gradually refines. Second, intentional temporal prediction practice. Tapping to a metronome refines internal time interval representation, improving temporal prediction accuracy. Third, utilizing visual cues. Attending to test screen layout and subtle pre-stimulus visual changes (screen flicker, frame update timing) can provide milliseconds of predictive advantage. Fourth, stabilizing internal clock through rhythmic breathing. Constant-rhythm breathing stabilizes cerebellar time processing, reducing temporal prediction variability. These approaches have immediate effects; conscious practice improves both reaction time stability and speed within 1-2 weeks.