GlossaryΒΆ
Glossary of all experiment attributes with short definitions.
- Blocking
Blocking an experiment means that the runs are grouped according to one or more factors, usually not controlled by the experimenters, such that all the runs are not randomized. The experiment is then separated in blocks, defined by the blocking factors.
- Center runs
Exprimental runs where all factors are set at the intermediate level (coded as 0). They are usually added in a design to provide an unbiased estimate of the variance or to perform lack-of-fit tests. Know aliases: center points, controls runs.
- DSD
To be done
- Foldover
To be done
- Fractional factorial
A design that is the fraction of a full-factorial design. A full-factorial \(2^{k}\) contains all combinations of the levels of \(k\) factors.
- Full factorial
A full-factorial \(2^{k}\) contains all combinations of the levels of \(k\) factors.
- Plackett-Burman
Plackett-Burman (PB) designs are saturated designs for two-level factors where the run size is a multiple of four. However, when the run size is also a power of two, the design is equivalent to a Fractional factorial design. Therefore, it is mainly used when the run size is a multiple of four, but not a power of two. In PB designs, for any two factors, all combinations of levels appear the same number of times.
- RCBD
Randomized Complete Block Design, or RCBD, is a blocked design (see blocking ), where each treatment appears exactly once in each block. It is commonly used in argicultural field trials.
- Saturated
A design is said to be saturated when there are \(k\) factors and \(k+1\) runs.
- Screening
To be done