Replication in Factorial Designs
Replication means repeating the same factor-level combination more than once.
In factorial designs, replication provides an estimate of experimental error, which allows standard errors, $t$-tests, $p$-values, and confidence intervals to be computed for factorial effects.