Three Quarter Plackett-Burman Designs for Estimating All Main Effects and Two-Factor Interactions
Plackett-Burman designs and three quarter fractional factorial designs are both well established in the statistical literature yet have never been combined and studied. Plackett-Burman designs are often non-regular and are thus subject to complex aliasing. However, Plackett-Burman designs have the a...
| Main Author: | |
|---|---|
| Format: | Others |
| Published: |
VCU Scholars Compass
2011
|
| Subjects: | |
| Online Access: | http://scholarscompass.vcu.edu/etd/2490 http://scholarscompass.vcu.edu/cgi/viewcontent.cgi?article=3489&context=etd |
| Summary: | Plackett-Burman designs and three quarter fractional factorial designs are both well established in the statistical literature yet have never been combined and studied. Plackett-Burman designs are often non-regular and are thus subject to complex aliasing. However, Plackett-Burman designs have the advantage of run-size efficiency (over the usual 2^(k) factorials) and taking three quarters of a Plackett-Burman design further improves this benefit. By considering projections of these designs, we constructed a catalog of designs of resolution V and ranked by D-efficiency. |
|---|