The apriori statistical power estimate depends on the ratio of the minimum biological effect to the coefficient of variation. Higher the ratio, greater the biological signal in a sample relative to noise. So, increasing the minimum biological effect value or decreasing the coefficient of variation value yields higher statistical power.
The formula to calculate the statistical power estimate is:
Power=Φ((μ-μ0)/(σ/√n) - z(1-α) )
where n=(σ (z(1-β) + z(1-α))/(μ-μ0 ))2
Φ is the cumulative distribution function of the standard normal distribution.
z(1-β) and z(1-α) derive from the quantile function of the standard normal distribution.
The minimum biological effect can be considered an expected signal for one or more analytes in a sample set, and the coefficient of variation can be considered as the noise. When the signal is greater than the noise, the statistical power is higher.
For practical reasons, the generally accepted p-value cutoff of 5% is used.
Steps
- Click Settings.
- Click Statistical Review.
- Do one of the following:
- Type a new value for Minimum Biological Effect.
- Type a new value for Coefficient of Variation.
- (Optional) Click Restore to system default to restore both assumption values to system default.