main effect

In an ANOVA the main effect is the measured effect of one of the conditions averaging out the contribution of all other factors. This is in contrast to the interaction effect between conditions.

For example, suppose we have measured the following error rates (averaged over many replications, and assuming equal numbers of replication in each cell):

	system A	system B
novice	70	50
expert	10	30

The main effect of expertise is that novices have (on average) 40 more errors than experts. The main effect of the system is zero as the decrease in performance for experts precisely matches the improvement for novices.