Developer Tutorial: New Capability to Simulate DOE Responses

This session is for JMP users who want to effectively create high-quality designed experiments.

A key benefit of simulating the response when designing your experiment is that you can be confident that the data will support your modeling and testing goals before conducting actual experiments. Most JMP DOE options allow users to manually enter coefficients in the existing Simulate Responses tool. The popular Random Coefficients option in JMP Easy DOE was strictly random, setting a few coefficients to small, non-zero values. It did not consider the effect type, or support the screening principes of effect sparsity, effect hierarchy, or model heredity.

Enrichments to the DOE Simulate Responses tool offer users the ability to simulate responses generated by random coefficients based on statistical principles.  This enables users to simulate responses quickly based on sound principles that model variances as well as means.

JMP Analytics Software Test Engineers Mark Bailey and Jacob Rhyne demonstrate and explain the capability and the underlying statistical approaches it deploys. The session includes time for Q&A.

This JMP Developer Tutorial covers: using JMP Easy DOE and Custom Design to initialize coefficients, including setting random seed, enforcing effect sparsity and enforcing effect hierarchy; and simulating the response.