Abstract
An in-depth analysis of Dorian Shainin's variable search (VS) method is carried out. Conditions under which the VS approach does and does not work well in comparison with traditional designs are identified. Explicit expressions for the expected number of runs and the probability of correct screening are derived under stated assumptions. The crucial roles of process knowledge and noise variation in successful application of the VS design are established through theoretical and simulation studies.
Additional information
Notes on contributors
Tirthankar Dasgupta
Dr. Dasgupta is an Assistant Professor in the Department of Statistics. His email address is [email protected].
Nagesh Adiga
Mr. Adiga is a Ph.D. Candidate in the School of Industrial and Systems Engineering. His email address is [email protected].
C. F. Jeff Wu
Dr. Wu is Professor of Engineering Statistics and Coca-Cola Chair in the School of Industrial and Systems Engineering. His email address is [email protected].