No abstract

We built an automated dialogue system whose style of interaction can be varied along the three dimensions of Humour, Relationship Maintenance and Personality Matching. We then ran a longitudinal experiment which investigated manipulations of these three dimensions. We explored the interaction of these separate dimensions on user perception of the system using a controlled study design. We showed a strong positive effect for the use of Humour and Relationship Maintenance, while the use of Personality Matching raised a number of questions which need further investigation.

The authors compared user evaluation of a text based dialogue system with a simple pick list for the same task. The authors matched the systems in terms of key factors, such as design for Usability, and took into account individual differences between participants that might have influenced their perception, such as Locus of Control, Experience and Personality. They found participants rated the text based dialogue system as being more informative, more credible, less frustrating and more persuasive than the simple pick list system. Participants’ ratings were dependent on their Personality, Locus of Control and reported level of physical activity. Participants did not differentiate between the systems in terms of their ease of use, indicating the other observed differences were not due to a simple difference in Usability. This study demonstrated the benefits of including automated dialogue in a system designed to help people find solutions for their exercise barriers. Further work is required to establish in what other situations dialogue provides a benefit.

The authors compared user evaluation of a text based dialogue system with a simple pick list for the same task. The authors matched the systems in terms of key factors, such as design for Usability, and took into account individual differences between participants that might have influenced their perception, such as Locus of Control, Experience and Personality. They found participants rated the text based dialogue system as being more informative, more credible, less frustrating and more persuasive than the simple pick list system. Participants’ ratings were dependent on their Personality, Locus of Control and reported level of physical activity. Participants did not differentiate between the systems in terms of their ease of use, indicating the other observed differences were not due to a simple difference in Usability. This study demonstrated the benefits of including automated dialogue in a system designed to help people find solutions for their exercise barriers. Further work is required to establish in what other situations dialogue provides a benefit.