Abstract
The "generation effect" is a phenomenon whereby people have better memory for information that is self-generated compared to information that is passively read. Throughout the years many theories have been proposed to explain this effect, one of which is the "mental effort theory," which suggests that more mental effort is allocated to self-generated information, meaning that the act of generating information inherently requires more mental effort than processing existing information. In a series of four paired-associates memory experiments, pupillometry (an independent measure of effort) was used to investigate a mental effort explanation of the generation effect within-subjects, between-subjects, and in a third experiment, within-subjects while manipulating generation difficulty. In a fourth, follow-up experiment, a verbal component was added to draw a link between generation quality and the pupillary response. All four experiments showed that more mental effort was allocated to generated information compared to read information, and that this was accompanied by a boost in memory performance when performed within-subjects. Importantly, in a cross-experimental covariance analysis for all within-subjects experiments, we found that differential effort allocation partially accounts for the behavioral generation effect. Taken together, the pupillometry results lend support to the idea that a mental effort is associated with the generation effect.