Contrasts

Contrasts are specific comparisons among treatments, chosen to address a scientific hypothesis. These can be represented as equations, like T1 - T2 = 0 which tests if T1 and T2 are equal. The coefficients of this example contrast are +1 and -1.

Single-degree-of-freedom contrasts ask single questions, like the example above.

A priori contrasts are chosen before the experiment is conducted and data examined.

Orthogonal polynomials are contrasts that examine polynomial trends among treatment means.