CONSTRAINED DESIGNS. PART I. FIRST ORDER DESIGNS.
Abstract
A common problem in industrial research is determining a cause and effect relationship between one or more quantitative variables, such as weight, volume or length, and some response such as performance or yield. In many problems the 'true' functional form relating the variables to the response is unknown. However, we can often 'graduate' this function in some particular region of interest by a polynomial or other flexible interpolation function. The task of the experimenter then becomes that of checking the adequacy of the graduating function, estimating the values of its coefficients and obtaining estimates of their precision and that of the fitted response.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1966
- Accession Number
- AD0647417
Entities
People
- C. J. Gardner
- George E. P. Box
Organizations
- University of Wisconsin–Madison