A two process account of the memory improvement due to choosing

Previous accounts of the mechanism which generates improvement in memory when people freely choose items, compared to other methods of item assignment, conflict and lack integration. I examine facial recognition performance of 40 participants asked to choose either the most or least attractive face from a series of pairs, and find that recognition of chosen faces is greater than recognition of unchosen, while no effect is found for the valence of faces. Considering these results in tandem with prior results and theories, I argue that a two process account of memory improvements due to choice is necessary, with one deliberative process occurring across all options available to choose from, and the other selective process focused on the actual chosen item. I detail the delineation of these processes, and describe and test the current best accounts of each — the multiple-cue hypothesis, and the self-reference effect for memory.