Definition and Calculation
The coefficient of variation (CV) is a statistical measure that evaluates data dispersion relative to the mean, calculated as standard deviation divided by the arithmetic mean, multiplied by 100 to express as a percentage (CV = SD / Mean x 100%). For a reaction time test where someone averages 200ms with a standard deviation of 30ms, the CV is 15%. The key advantage of CV is enabling fair comparison of variability between datasets with different means. A person averaging 200ms and another averaging 300ms with the same 30ms standard deviation have different relative variability, and CV correctly captures this distinction.
What CV Reveals in Cognitive Testing
Reaction time CV serves as an important indicator of cognitive consistency. Healthy adults typically show relatively small reaction time CVs, though the usual range depends on the task. High CV (large variability) suggests unstable attention maintenance and may serve as an early marker of attention deficits or fatigue. Research has repeatedly reported that individuals with ADHD have higher reaction time CVs than controls. CV also tends to increase with aging, a change that mean reaction time slowing alone does not explain and that is thought to reflect declining stability of the nervous system. This means CV can add information that average speed alone does not provide.
Practical Use in Bench Tests and Improvement Strategies
Variability in reaction time carries information that the mean alone does not. A fast average combined with large trial-to-trial differences may indicate the presence of extremely slow trials (lapses), suggesting attention sustainability challenges. The Bench result screen shows the spread across your trials, so it can be read alongside the average as a cue to stability. The primary improvement strategy is ensuring adequate sleep quality and duration - sleep deprivation has been reported to affect CV more than mean reaction time. During testing, maintaining a consistent rhythm and minimizing both extremely fast and extremely slow trials works to lower CV. A decrease in CV shows that your responses have become more stable, which is a different kind of gain from a faster average.