Factor

An analysis of variance using preselected factor’s levels from the range of values of some factors, and some factors have randomly selected levels from the corresponding factor’s assumed distribution.

A graphical design of experiments analysis method plotting the main effects plot for a single factor constructed or each level of another factor. Parallel lines indicate little or no interaction between factors. When not parallel, indicates factor interaction. Very different slopes indicate high degree of interaction.

The non-straight-line behavior between one or more factors and the response, usually expressed in mathematical terms involving the square or cube of a factor. Also defined as a departed from a straight line between the response and predictor variables.

A basic statistical technique for determining the proportion of influence a factor or set of factors has on total variation. It subdivides the total variation of a data set into meaningful component parts associated with specific sources of variation to test a hypothesis on the parameters of the model or to estimate variance components. There are three models: fixed, random, and mixed. ANOVA provides a means to statistically analyze two or more factors at a time.

The arrangement in which an experimental program is to be conducted and selection of the levels of one or more factors or factor combinations to be included in the experiment. It is also known as statistical design of experiment, as it entails a versatile statistical approach to screen for important factors, optimize factor settings, or detect important factor relationships or interactions.