TREATMENTS factor-selections;
The TREATMENTS statement specifies the treatments of the plan to generate, but it does not generate a plan. If you supply several FACTORS and TREATMENTS statements before the first RUN statement, the procedure uses only the last TREATMENTS specification and applies it to the plans that are generated by each of the FACTORS statements. The TREATMENTS statement must follow a FACTORS statement. The TREATMENTS statement has the same syntax as the FACTORS statement. The individual factor-selections also have the same form as in the FACTORS statement:
name = m <OF n> <selection-type>;
The procedure generates each treatment simultaneously with the lowest (that is, the most nested) factor in the last FACTORS statement. The m value for each treatment must be at least as large as the m for the most nested factor.
The following statements use both a FACTORS
and a TREATMENTS statement. First the FACTORS
statement sets up the rows and columns of a square (factors r
and c
). Then, the TREATMENTS statement augments the square with two cyclic treatments.
The resulting design is a Graeco-Latin square, a type of design useful in main-effects factorial experiments.
proc plan; factors r=3 ordered c=3 ordered; treatments a=3 cyclic b=3 cyclic 2; run;
The resulting Graeco-Latin square design is shown in FigureĀ 86.5. Notice how the values of r
and c
are ordered (1, 2, 3) as requested.
Figure 86.5: A Graeco-Latin Square