The memory task used in this study involved teaching subjects the correct location of 50 different icons on a computer screen. Icons - small pictures of a cat or a tea kettle - were associated with relevant sounds (for example, a "meow" sound was played when the picture of the cat was on the screen). The subjects then took a nap (less than 90 minutes of sleep), during which they were monitored via EEG to track their sleep stage. During the nap, some subjects heard only white noise, while some other subjects were (unknowingly) exposed, during "slow-wave" sleep, to 25 of the sounds they'd heard during the picture location learning task. After the nap, subjects were tested for their ability to place icons on the screen.
Subjects who, during their nap, had heard the sounds associated with the icons performed significantly better in the memory task than subjects who heard only white noise during the nap. EEG patterns were measured between the two groups, and confirmed that there was a significant difference between the electrical activity of their brains. Though not exhaustive, the researchers performed a variety of control experiments to determine whether or not the sounds played during slow-wave sleep were the causal factor in the improved performance of the subjects who heard them, and it seems reasonable to infer that they were.
I would like to see some continued research on the importance of icon-sound relevance relative to task performance post-nap. In other words, would performance on the task be as improved if subjects had heard an "irrelevant" sound for each icon - say, a car horn honking when the icon of the cat was displayed? Of course, it would also be great to see further research on different types of memory and learning tasks and potential improvements in performance by associating sensory experiences (sound, smells) during the learning process and during sleep following that learning experience. I am also curious about the utility of listening to music while studying, and then listening to that same music while napping. In the mean-time, so long as the music we choose doesn't distract from learning or from sleep, it would seem reasonable to suggest that memory task performance may increase as a result.