What is One factor at a time (OFAT)?
One factor at a time (OFAT) is an effective problem-solving technique that identifies the most critical causes of an effect. Ceteris paribus, the approach is to change only one factor. All other factors (causes) are kept constant. Hypothesis testing is one of the most common tools used for One Factor at a Time (OFAT).
Non-experts prefer One Factor at a Time (OFAT), particularly in situations where data is abundant and cheap.
One Factor at a Time (OFAT) can be used in cases where the mental work required to perform a multi-factor analysis is greater than the time required to collect additional data. Researchers have also shown that One Factor at a Time (OFAT) is more effective than partial factorials in certain circumstances (e.g., the number of runs is limited, the primary goal is system improvement, and experimental errors are not large when compared with factor effects which must be independent and additive).
It is better to alter multiple factors simultaneously in situations that require careful analysis of data. This point can be illustrated by the family of Balance Puzzles which includes the Twelve Coins. On the undergraduate level, one can compare Bevington’s
GRADLS. The former is not optimal, while the latter, which only changes one variable at a given time, is even worse. See also the factorial experiment design pioneered by Sir Ronald A. Fisher. OFAT is not liked for a number of reasons.
- OFAT requires multiple runs to achieve the same accuracy in effect estimation
- OFAT does not estimate interactions
- OFAT may not be able to set factors at their optimal setting.
The Plackett Burman is a modern design that has all factors varied simultaneously, which is an important feature in experimental designs, and gives greater accuracy in effect estimation.