with respect to statistics, a moderator variable is entirely unrelated to the criterion variable but it is still related to the other predictor variables which means it will be included in the regression equation. There are other applications of a moderator variable, for example in the context of consumer behaviours.

MODERATOR VARIABLE: "Moderator variables are unrelated to any criterion values which are used in an experiment, conversely however they are still related to any other moderator variables."