Random coding values

[#1]

Sometimes when setting up a DoE in JMP, the +1, -1 coding  is joined by seemingly random additional values, as in the -0.3 value in the snip below. I assume such a value is genuine and not a mistake.  Can anyone explain what such a value means?

[#2 · Accepted solution]

Hi @kjwx109prime,

The value you're seeing is the result of the convergence of the Coordinate-Exchange algorithm used to compute optimal design in the platform Custom design, based on your assumed model, optimality criterion and any constraints you may have added on the design space. More info on the algorithm : https://www.linkedin.com/feed/update/urn:li:activity:7163071856332234752/ 

If there are no constraints in your design and you obtain this type of value, you can try to increase the design search time and/or the number of random starts in the Custom design platform options to try to obtain more homogeneous values in your design. 

See closely related posts and answers here :

How are odd factor settings in D-optimal RSM generated 

Random decimals incorporated in mixture screening design 

Hope this response will help you,