The design of experiments (DOE)
The design of experiments (DOE, DOX, or experimental design) is the design of any task that
aims to describe and explain the variation of information under conditions that are
hypothesized to reflect the variation. The term is generally associated with experiments in
which the experiment design introduces conditions that directly affect the variation. The
conditions that influence the variation are selected for observation.
In its simplest form, an experiment aims at predicting the outcome by introducing a change
of the preconditions, which is represented by one or more independent variables, also
referred to as "input variables" or "predictor variables." The change in one or more
independent variables is generally hypothesized to result in a change in one or
more dependent variables, also referred to as "output variables" or "response variables." The
experimental design may also identify control variables that must be held constant to prevent
external factors from affecting the results. Experimental design involves not only the
selection of suitable independent, dependent, and control variables, but planning the
delivery of the experiment under statistically optimal conditions given the constraints of
available resources. There are multiple approaches for determining the set of design points
(unique combinations of the settings of the independent variables) to be used in the
experiment.
Main concerns in experimental design include the establishment of validity, reliability,
and replicability
Orthogonal arrays (OA)
Orthogonal arrays (OA) represent a simplified method for an experiment. The original
development of the concept was by Sir R.A. Fischer of England in 1930s.
Taguchi constructed a special set of orthogonal arrays. His approach takes each of the
factors at two levels (usually, the extremes) and works out which has the greatest
contribution to the end result. These factors are then studied in more details. An orthogonal
array is shown in table below. The 8 in the designation OA8, represents the number of rows,
which is also the number of treatment conditions. Each row thus, represents a trial
condition with factor levels indicated by the numbers in the row. Across the top of the
orthogonal array is the maximum number of factors that can be used; which in this case is
seven (A to G).
Table 1: Orthogonal Arrays
Trial Number (Tc ) A B C D E F G
1 1 1 1 1 1 1 1
2 1 1 1 2 2 2 2
3 1 2 2 1 1 2 2
4 1 2 2 2 2 1 1
5 2 1 2 1 2 1 2
6 2 1 2 2 1 2 1
7 2 2 1 1 2 2 1
8 2 2 1 2 1 1 2
The levels are assigned by 1 and 2. If more levels occur in the array then 3, 4, 5 and so forth
are used. The orthogonal property of the OA is not compromised by changing the rows or
the columns. The array forces all experiments to design almost identical experiments. Thus
it assures consistency of design by different experiments.
The stages to go through are:
1. Selection of the factors to be evaluated.
2. Selection of the number of levels.
3. Selection of the appropriate orthogonal arrays.
4. Assignment of factors to the columns.
5. Conduct the experiment.
6. Analyse Data
7. Carry out the experiment
Taguchi method gives a fast and pragmatic approach to the optimization of products and
processes. It can also be used for tolerance design in setting of statistically based tolerances,
allowing either improved performance through tighter tolerances or cheaper designs in non-
critical areas.
Degrees of Freedom
The number of degrees of freedom is very important value because it determines the
minimum number of treatment conditions. It is equal to:
Number of factors (Number of levels -1) + Number of interactions (Number of levels -1)
(Number of levels -1) + one for the average.
The following example will illustrate this concept:
Example: Given four two-level factors, A, B, C, D and two assumed interactions, BC and CD,
determine the degrees of freedom, d f
i. For two-level factors
d f= 4(2-1) +2(2-1) (2-1) +1=7.
Therefore, at least seven treatment conditions are needed for the two-level.
What is the answer if the factors are three-level?
ii. For three-level factor:
d f= 4(3-1) +2(3-1) (3-1) + 1 =17.
Thus, at least seventeen treatment conditions are needed for the three-level.
It is therefore, clear that the number of levels has considerable influence on the number of
treatment conditions. Although a three-level design provides a great deal more information
about the process, it can be costly in terms of the number of treatment conditions.
The maximum degrees of freedom is equal to:
f
l
Where l = number of levels
f = number of factors
For the example problem with two levels and 4 factors
4
l = 2 =16
f
Table 2. below show the maximum degrees of freedom.
Table 2: Maximum degrees of freedom for a four-factor, two-level experimental design:
Design space df
A B C D 4
AB AC AD BC BD CD 6
ABC ABD ACD BCD 4
ABCD 1
Average 1
Sum 16
