| Multiple resource theory suggests that enhancements in human information management capacity may be realized via multimodal interaction. Leveraging multiple sensory systems to maximize working memory (WM) throughput becomes essential as the information age conveys volumes of data that would overburden the visual channel alone. The current study proposed an expansion of the current bimodal (verbal, visual/spatial) model of WM to a multimodal WM system, which includes verbal, visual, spatial, kinesthetic, tactile, and tonal component subsystems. Single modality capacity was measured for each proposed subsystem. In addition, multimodal capacity was calculated for combined modalities. Experiments 1 and 2 demonstrated that multimodal WM capacity surpasses that of single modality capacity. Most notably, multimodal WM capacity averaged more than three times the 'magic number' seven proposed by Miller's (1956) unidimensional memory span. Results in Experiment 2 also indicated that multimodal capacity nearly reached the summation of each single modality capacity. A dual-task paradigm was used to explore the independence of the proposed WM subsystems. Results in Experiment 3 demonstrated minimum interference between the subsystems in terms of storage and rehearsal mechanisms; which supports the proposed multimodal WM model. Theoretical and applied implications for the use of multimodal interaction are discussed. A set of preliminary design guidelines for multimodal information systems is suggested. |
