Measuring learning using an untrained control group: comment on R. Reber and Perruchet.

R. Reber and Perruchet (this issue) argue that use of control groups without training is unsound for establishing that learning has occurred. We show that inferring learning from a difference between a trained group and an untrained control in no way relies on their implausible additivity assumption, and that untrained control groups can be an invaluable aid to the researcher.