Interaktionsplots

Interaktionsplots, in statistics and data visualization, are plots used to visualize interaction effects between two or more factors in factorial experiments. They display the response means for combinations of factor levels, with separate lines for the levels of one factor and the other factor on the axes, or with panels for one factor.

Construction and interpretation: in a two-factor design, compute the mean response for each level combination, plot the mean on the vertical axis against the level of one factor on the horizontal axis, and use distinct lines for the levels of the second factor. If the lines are roughly parallel, there is little or no interaction between the factors; if the lines diverge, converge, or cross, this suggests an interaction where the effect of one factor depends on the level of the other.

Relation to statistics and use: interaction plots are often used after fitting an ANOVA or a regression model to visualize the interaction term and guide interpretation. They are also useful as exploratory tools to identify potential interactions and to plan follow-up analyses, such as simple effects tests or estimated marginal means plots.

Limitations and considerations: plots can be sensitive to scale, unequal group sizes, or outliers and may mislead when relationships are nonlinear or when higher-order interactions are present. They should be interpreted cautiously and complemented with formal tests. Extensions include plotting estimated marginal means or using facet plots to display more factors, while keeping interpretability in view. Applications span psychology, agronomy, engineering, and other fields that employ factorial designs.