Image memorability depends on interference in memory

Some experiences are better remembered than others by most people. This inherent characteristic of an event to be well remembered is known as memorability. Memorability has been typically measured by performance on recognition memory tasks and is calculated as the proportion of participants who correctly recognize a repeated item. However, it is unclear what drives certain experiences to be more memorable than others and whether all aspects of an experience are better remembered. Lure discrimination can capture more nuanced aspects of our memories by measuring our ability to discern similar “lure” items that share overlapping features with items encountered previously. Interference across experiences is a major barrier that our memory system must overcome, yet we do not know how interference may impact the memorability of our experiences. Exploring this facet of memory can allow us to parse what drives certain experiences to be memorable or forgettable. To this end, we developed a memory task that varied image memorability and lure image similarity and measured both target recognition - memory for repeated images, and lure discrimination - discrimination of similar lure images. While effects of memorability were evident for target recognition, the impact of memorability on lure discrimination depended on lure similarity, where higher levels of interference did not benefit from memorability compared to lower levels of interference. Furthermore, these effects were time-dependent: memorability impacted memory immediately, but not after longer time delays. These findings suggest that the concept of memorability depends on how and when memory is measured.