Abstract
Supersaturated designs are designed to assess the effects of many factors simultaneously. The assumption of "effect sparsity" is often needed to justify the selection of these designs. However, when effect sparsity holds, Type I errors can easily occur. Forward-selection multiple test procedures are proposed to address and solve this problem.
Original language | English (US) |
---|---|
Pages (from-to) | 101-117 |
Number of pages | 17 |
Journal | Statistica Sinica |
Volume | 8 |
Issue number | 1 |
State | Published - Jan 1 1998 |
All Science Journal Classification (ASJC) codes
- Statistics and Probability
- Statistics, Probability and Uncertainty