UCLA

A common desire in the modern world is to improve how much we remember about key daily events, and improving this requires understanding how information is processed in memory. One promising yet often overlooked method may be to utilize sensory integration. Previous work performed by multisensory research groups have shown that multisensory stimulus presentation can improve memory performance for facts and episodes. However, whether these findings are general and robust, what kind of tasks could benefit from multisensory encoding, and what the underlying mechanisms are questions still unanswered. Some of these limitations emerge from the limited number of studies investigating memory through a multisensory framework, and that these studies do not always replicate one another’s results. Other limitations come from the interplay of these multisensory studies with existing memory theories, almost all of which do not acknowledge that sensory combination and sensory integration are distinct. Thus, I sought to answer remaining key questions in the field of multisensory memory encoding. First, with discrepant findings amongst multisensory research groups regarding the presence of memory benefit, I investigated drift-diffusion modeling and other simultaneous measures of speed and accuracy as tools to quantify multisensory benefit (Study 1). This discovered that such measures were a sensitive and reliable measure of multisensory benefit, which was later applied to investigate if speed-accuracy tradeoffs were present in our empirical memory findings. Second, as multisensory memory benefit has only previously been explored in a limited variety of memory tasks for basic objects, we sought to expand the body of research to include more challenging associative memory tasks, specifically memory for face-name associations (Study 2) and Swahili-to-English vocabulary memorization (Study 3). Study 2 showed that multisensory stimulus presentation, specifically, is helpful for bolstering associative memory for faces and names. Study 3 provides an interesting case where multisensory presentation fails to produce better memory, providing insight to important border conditions that have not been previously discussed. Finally, as existing memory theories do not separate multisensory processes from the mere presence of information across senses, I investigated whether multisensory representations are stored in memory, and whether this is true for all individuals (Study 4). By testing participants’ memory for an illusion, it was shown that, for the vast majority of participants, multisensory representations are coded, meaning most participants should specifically benefit in their memory performance from encoding information in a multisensory manner. These studies shed light on the mechanisms of encoding and retrieval in ecological multisensory experiences and may have translational implications for facilitation of memory in everyday tasks.