foldover(object)
"design"
.
"design"
which is the original
design folded over, i.e., a design with twice the number of rows than
the original design.
The first half of the returned design is identical to the original design, the second half is the first half with the levels reversed (i.e. the first level of each factor substituted for the second, and vice versa).
A foldover design is created by taking each of the runs in the
original design and reversing the signs (levels). A foldover design is
typically used to "unconfound" a Resolution III design; that is, if
you fold over a resolution III design, it becomes a resolution IV
design.
This can be useful when the results from a first small
experiment indicate that there are possibly many interactions which
can then be separated from the main effects by running the second half
of the folded over design.
A foldover of a twelve-run
Plackett-Burman design, is a Resolution IV design. The resulting
design has the desirable property that all two-factors interactions
with the first factor are estimable (although these two-factor
interactions are confounded with other two-factor interactions).
Such designs can be useful alternatives to 32-run designs.
Box, G.E.P., Hunter, W.G., and Hunter, J.S. (1978).
Statistics for Experimenters. New York: Wiley.
Box, G.E.P. and Wilson, K.B. (1951). On the experimental attainment of
optimum conditions. J. Royal Statistical Society, Series B, 13. 1.
Haaland, P. D. (1989).
Experimental Design in Biotechnology.
New York: Marcel Dekker.
# foldover a 7 factor 8 run design to get a resolution III # design alias(design.digest('ff0708')) alias(foldover(design.digest('ff0708'))) # create a foldover of a Placket-Burman design pb12.des <- design.digest('pb1112') alias(pb12.des) pb12.des.fo <- foldover(pb12.des) alias(pb12.des.fo)