Statistical Design of Experiments
Statistical Design of Experiments
Using 3 levels allows for estimation of a
Using 3 levels allows for estimation of a
quadratic relationship between factor and
quadratic relationship between factor and
response
response
In this example, to minimize the response,
In this example, to minimize the response,
the user would want to know the effect of
the user would want to know the effect of
the medium setting
the medium setting
Consider the same exact factor,
Consider the same exact factor,
only experimented at a third
only experimented at a third
level between Low & High
level between Low & High
i.e. same data for Low & High,
i.e. same data for Low & High,
just an additional point in
just an additional point in
between
between
Factor X
Factor X
3 Level Factor
3 Level Factor
10
10
6
6
8
8
2
2
4
4
Low
Low
High
High
Medium
Medium