Definition
A statistically planned method for deliberately varying controlled input factors so their effects and interactions on measured responses can be estimated efficiently and rigorously.
Application
Used in semiconductor process development and optimization to identify significant factors, build response models, determine robust operating windows and learn interactions using fewer, more informative experiments than one-factor-at-a-time trials.
Related search terms
Designed experiments; experimental design
Search phrases help you find this entry; their meaning depends on the context shown here.
Related concepts
Evidence and further reading
- Evidence from itl.nist.govhttps://www.itl.nist.gov/div898/handbook/pri/section1/pri11.htm
- Evidence from itl.nist.govhttps://itl.nist.gov/div898/handbook/pmd/section3/pmd31.htm