Work out time study sample size instantly with clear inputs, formula shown and shareable results.
For a continuous variable such as cycle time, sample size is n = (z x coefficient of variation / relative accuracy)^2. Variability drives everything: a task with a 10% coefficient of variation needs about 16 cycles for 5% accuracy at 95% confidence, while one at 30% needs around 139.
Time study sample size
n = (z x s / (accuracy x mean))^2
Sample size grows with the square of the coefficient of variation. Rather than observing more, it is usually better to find and remove the source of variation — inconsistent method, material or machine condition.
Only if the cause is identified and genuinely non-recurring, such as a tool break. Excluding slow cycles because they are inconvenient produces a standard nobody can achieve.