Main Effects and Interaction Analysis: Visualize DOE Effects
What is a Main Effect?
A main effect is the change in the average response when a factor moves from its low level to its high level. The main effects plot connects the mean response at each factor level; the steeper the line, the stronger the factor's influence, while a nearly flat line indicates little effect. Main effects capture each factor's isolated contribution and are usually the first output examined in a DOE analysis. Effects are typically expressed in coded units (-1 to +1) so that factors can be compared directly on the same scale.
What Is an Interaction?
Two factors interact when the effect of one factor depends on the level of the other. In an interaction plot, parallel lines indicate no interaction, while crossing or clearly non-parallel lines signal a significant interaction. When a strong interaction exists, interpreting main effects alone can be misleading; you should examine simple effects by looking at one factor within each level of the other. Factors that interact should be optimized together rather than tuned independently.
Pareto Effect Plot
The Pareto effect plot sorts the absolute values of all main and interaction effects as bars and overlays a significance reference line based on a t-test or Lenth's method. Effects above the line enter the model, while those below are pooled into error, helping you trim the model and focus on the factors that matter. The reference line adjusts to the robustness of the effect estimates and is especially useful when there are no replicated runs.
How to Use It (Step by Step)
Log in and enter your factor-level combinations and response data in DOE table format, or import a file generated by the DOE tool. The tool calculates each effect, produces main effects, interaction and Pareto effect plots plus an ANOVA table, and ranks factors by effect size with recommended optimal level settings. You can enable Lenth's method for robust significance testing when the dataset is small.