Multisensory Facilitation of Working Memory Training

Abstract

Research suggests that memorization of multisensory stimuli benefits performance compared to memorization of unisensory stimuli; however, little is known about multisensory facilitation in the context of working memory (WM) training and transfer. To investigate this, 240 adults were randomly assigned to an N-back training task that consisted of visual-only stimuli, alternating visual and auditory blocks, or audio-visual (multisensory) stimuli, or to a passive control group. Participants in the active groups completed 13 sessions of N-back training (6.7 hours in total) and all groups completed a battery of WM tasks: untrained N-back tasks, Corsi Blocks, Sequencing, and Symmetry Span. The Multisensory group showed similar training N-level gain compared to the Visual Only group, and both of these groups outperformed the Alternating group on the training task. As expected, all three active groups significantly improved on untrained visual N-back tasks compared to the Control group. In contrast, the Multisensory group showed significantly greater gains on the Symmetry Span task and to a certain extent on the Sequencing task compared to other groups. These results tentatively suggest that incorporating multisensory objects in a WM training protocol can benefit performance on the training task and potentially facilitate transfer to complex WM span tasks.

Keywords: multisensory; training; transfer; working memory.

Figure 1:

Enrollment

Figure 2:

Timeline of training and testing for passive and active groups. A 5–7 minute break was given between testing and training blocks and between two 20-minute training blocks.

Figure 3:

Mean N-level achieved on a given day, split by visual and auditory blocks for the Alternating group. A dip on day 8 reflects transition from less abstract stimuli to more abstract stimuli (equivalent across groups).

Figure 4:

Difference scores for each group: V = Visual Only, A = Alternating Audio-visual, M = Multisensory, and C = Control (Passive); color coded as red, blue, green and gray, respectively. (A) Difference between highest N-level achieved at post-test relative to pre-test on an untrained visual N-back task. (B-D) Change in span on Forward Corsi Blocks, Sequencing, and Symmetry Span, respectively. Error bars represent S.E.M. * p <0.05, ** p < 0.01. Stars without lines indicate significant difference at post-test relative to pre-test in each group (Wilcoxon signed-rank test), whereas stars with lines indicate significant differences in change scores in one group relative to another (Kruskal-Wallis test).

Figure 5:

Self-reported improvement on tasks at post-test as a result of earlier sessions, shown as percentage per group; V = Visual Only, A = Alternating Audio-visual, M = Multisensory, and C = Control (Passive).