Multiple determination is an axiom of life sciences. Behavior cannot be understood in terms of single variables. The classic Experimental–Control paradigm, basic though it is, misses the importance of multiple determination.
A conceptual discontinuity occurs between one and two variables. With two variables, some assumption about their integrated action becomes necessary. Each separate variable may have its own effects, but the observed response depends on their joint action. Each may modify the other's effects, moreover, so possible inter actions must be allowed. To extend experimental analysis to multiple determination thus requires new ideas. Such new ideas appear with factorial design.
Scientific Method.Factorial design, developed extensively by R. A. Fisher in conjunction with analysis of variance, represented a fundamental advance in scientific method. Factorial design has become a basic tool for experimental analysis of multiple determination.
Previously, as Fisher observed, exposition of scientific method had emphasized the rule of varying just one variable at a time. This rule seemed sensible because it ensured that any effect would have an unambiguous causal interpretation. As a consequence, however, multiple determination was typically studied in unsystematic, often haphazard ways.