Personal response systems, commonly called ‘clickers’, are widely used in secondary and post-secondary classrooms. Although many studies show they enhance learning, experimental findings are mixed, and methodological issues limit their conclusions. Moreover, prior work has not determined whether clickers affect cognitive change or simply alert students to information likely to be on tests. The present investigation used a highly controlled methodology that removed subject and item differences from the data to explore the effect of clicker questions on memory for targeted facts in a live classroom and to gain a window on the cognitive processes affecting the outcome. We found that in-class clicker questions given in a university psychology class augmented performance on delayed exam questions by 10–13%. Experimental results and a class survey indicate that it is unlikely that the observed effects can be attributed solely to attention grabbing. Rather, the data suggest the technology invokes the testing effect. Copyright © 2012 John Wiley & Sons, Ltd.