The Residual column was calculated as follows:
From the output below, you can see that the numbers in the formula above are the MLEs in a Weibull regression with linear length effect.
Now we look at how the exponential residual column was calculated:
If the Weibull regression model fits, this column should have a Weibull Distribution with slope one. We check this using a Weibull plot:
As we can see this model does not seem to fit the data well. Let's try to add a quadratic length effect to the Weibull regression.
With this information we calculate another residual column
This residuals indicate a better fit but we are not there yet.
Another residual plot of interest is the following:
This plot indicates that we have removed the effect of length from these residuals.
Find out how to download a JMP file
