There was also a significant difference in the six modes for Subject Preference, F(5,138) = 105.6, p > 0.0001. The pen + voice mode had the highest satisfaction rating (mean = 8.5, see Figure 2). And only-voice mode has the lowest satisfaction rating of 6.3.
Based on the analyses, the pen + voice combination is the best of the six interaction modes. This experiment confirmed the Hauptmarm et al. ( Hauptmann A. G. 1989) finding of surprising uniformity and simplicity in the user's gestures and speech, and Oviattet al. ( Oviatt, S., DeAngeli, A., and Kuhn, K. 1997) observation that users overwhelmingly preferred to interact multi-modally rather than single-modally. Nitta ( Nitta, T. 1995) also raises this point.
Adam Cheyer and Luc Julia ( 1996), Multimodal Maps: An Agent-based Approach; SRI International, 1996, http:// www.ai.sri.com/ ~cheyer/papers /mmap/ mmap.html.
Bekker, M. M., Nes, F. L.van, and Juola, J.F( 1995), A Comparison Mouse and Speech Input Control of A Text-annotation System, Behaviour & Information Technology, Vol. 14, No. 1, 14-22,
Hauptmann A. G. ( 1989), Speech and Gestures for Graphic Image Manipulation, In Proceeding of the CHI'89 Conference on Human Factors in Computing Systems, 241-245.
Nitta, T. ( 1995), From GUI towards Multimodal UI (MUD) (in Japanese), Information Processing Society of Japan, Vol. 36, No. 11, 1039-1046.
Oviatt ( 1996), Multimodal Interfaces for Dynamic Interactive Maps. In Proc. CHI '96, ( Vancouver), 95-102.
Oviatt, S., DeAngeli, A., and Kuhn, K. ( 1997), Integration and Synchronization of Input Modes During Multimodal Human-computer Interaction, In Proceeding of the CHI'97 Conference on Human Factors in Computing Systems, pp.415- 422, 1997.
Ren, X., and Moriya, S. ( 1997), The Strategy for Selecting a Minute Target and the Minute Maximum Value on a Pen-based Computer, In Extended Abstract of the CHI'97 Conference on Human Factors in Computing Systems, 369-370,