The Difference Between Mediators and Moderators
The main objective of inferential statistics is to examine relationships between variables. Usually, these relationships are linear, which means that when one of the variables changes, the other tends to change proportionately. Stronger or weaker relationships mean that as one changes, the other changes by a larger or smaller amount. This relationship can be positive or negative, which means that as one increases, the other should either increase or decrease, respectively. However, some types of relationships are more complicated, and are influenced by the presence of third variables. Examining them can often provide a clearer, more useful interpretation of the data. There are two main types of these third variables: moderators and mediators.
Moderators are variables which influence the strength or direction of a relationship. For example, suppose the use of practical word problems resulted in higher math test scores. However, the need to read and comprehend the word problems suggests that reading ability might also play a role. While reading ability is unlikely to be directly related to math test scores, it might be a moderator, such that higher reading ability would result in greater improvement, and lower reading ability would result in less improvement from the problems, or even reduced performance.
Unlike moderators, mediators determine whether a relationship exists at all. For example, suppose a relationship existed between motivation and school grades. However, motivation would only result in better performance in the presence of a third variable: time studying. Without study, motivation is unlikely to have a significant impact on grades, as motivation is only meaningful if you act on it.
- Moderators are variables that influence the strength or direction of the relationship between two other variables.
- Mediators control whether a relationship between two other variables exists.