Usability of embodied conversational agents in a multi modal mobile serious game

There are many theoretical advantages in favour of ECAs and spoken dialogue systems because they provide a more “natural interaction”. [1] Although, there is a growing pool of empirical data relevant to the effects of ECAs, there is still lack of empirical evaluations proving that ECAs are more usable on mobile devices. Given the lack of evidence on the potential effect of ECAs on serious games, there is a major risk related to the introduction of ECAs in serious game mobile applications. In fact, the ECA might not enhance the application or might not be appropriate in this context of use. This could lead to a poor performance by the player as the ECA might act as a distraction rather than a helpful element and the interaction may collapse. [2] Therefore, whether usability and quality are to be enhanced by the use of an ECA in a multimodal human-machine interface, has to be decided for each application anew. [3]

[1]Takeuchi, A, & Naito, T (1995). Situated facial displays: Towards social interaction, In Proc. Conf. on Human Factors in Computing Systems (pp. 450–455).

[2]I. Doumanis and S. Smith, 'A Framework for Research in Gamified Mobile Guide Applications using Embodied Conversational Agents (ECAs)', International Journal of Serious Games , vol. 2, no. 3, 2015.

[3]B. Weiss, I. Wechsung, C. Kuhnel and S. Moller, 'Evaluating embodied conversational agents in multimodal interfaces', Computational Cognitive Science , vol. 1, no. 1, 2015.