Interaction Plot Generator

For each of the below, enter the estimated coefficient from the model. The Intercept refers to the actual intercept/constant (or see below), X and M refer to the variable of interest and the moderator respectively, and the Interaction to their estimated interaction term.

Estimated Intercept:

Estimated Slope on X:

Estimated Slope on M:

Estimated Interaction:

Additional details regarding Intercept

If your model has no additional covariates, you can use the intercept/constant reported from the model as the Intercept. However, if you have covariates, you likely want to use an intercept at some "average" value of those covariates. One way to obtain this is to center (at their mean, or some specific meaningful value) each of the other covariates, then use the reported intercept/constant.

Alternatively, you can manually compute an intercept. Say your model has a single other predictor Z. Let Z* represent some "average" value of Z (again, this can be the actual mean, or a representative value; if Z was IQ, perhaps Z* being 100 would make sense, regardless of the actual sample mean of Z). Let α be the reported intercept, and let β be the estimated coefficient on Z. You would then use α + βZ* as the Intercept.

Enter at least 2 values for X. Only two points are required to draw a line in the linear case, though more can be added if desired.

Values for X:

For each value of the moderator here, a separate line will be drawn. You can enter values such as the mean and +/- 1 standard deviation, or you can choose several meaningful values (e.g. if the moderator is age, perhaps 21, 40 and 65 as legal age, middle age, retirement age). Adding too many values will make the plot very messy.

Values for Moderator:

Draw individual points?