The expected variation is equal to the standard deviation of the transformed counts (Xi), which is equal to 1.
To calculate the observed variation, Minitab calculates normal scores (Y,) for the transformed counts as follows:
where NSCOR is the Normal scores function (available by choosing ).
For the next step, only the middle 50% of the Xi values are used, along with their corresponding Yi values. Xi values are excluded if they are less than the 25th percentile or greater than the 75th percentile.
Minitab fits a least squares regression model with Yi as the response and Xi as the predictor:
The observed variation is then 1 / β1.
The ratio of observed variation to expected variation is calculated as follows:
|Xi||transformed counts (For more information, see the section "Plotted points".)|
|β0||intercept from the least squares regression equation|
|β1||slope coefficient from the least squares regression equation|