Well-conditioned models are important, particularly for practitioners who work with regression models for mixture experiments where parameter estimates are individually meaningful. In this article we ...