Abstract
We have developed an animated PDA-based advisor that can engage sedentary adults in dialogues about their physical activity throughout the day. An integrated accelerometer enables the advisor to initiate interactions and provide real-time feedback. Results of preliminary usability testing of interaction modalities are presented and a planned efficacy study using free-living sedentary adults is described.
Problem
Participation in moderate amounts of physical activity has important health benefits, including beneficial effects on risk factors for disease, disability, and mortality. Yet, a substantial proportion of the U.S. adult population remain underactive or sedentary.1
System Description
A PDA-based exercise advisor system comprises an animated character, a dialogue system and an integrated sensor (Figure 1). The advisor is capable of a range of nonverbal conversational behavior, including: facial displays of emotion; head nods; eye gaze movement; eyebrow raises; posture shifts and “visemes” (mouth shapes corresponding to phonemes). Advisor output utterances are displayed as text with the words individually highlighted at normal speaking speed and the nonverbal behavior displayed in synchrony. User inputs are constrained to multiple choice selections at the bottom of the display. Interaction dialogues are scripted in a custom general-purpose XML scripting language that specifies a state transition network. This system design parallels a desktop computer-based exercise advisor system that has been successfully evaluated in two randomized studies.2
Figure 1.
Exercise Advisor System
We are currently integrating an accelerometer into the system (shown at the bottom of Figure 1) so that it can detect when users are performing moderate-or-greater intensity walking (50% heart rate reserve, calibrated for each user). This information will enable the system to initiate dialogues with the user, for example, providing positive feedback immediately following a bout of walking or problem solving if the user is not actually walking at a time they had previously committed to go walking.
Evaluation & Acknowledgements
Preliminary usability testing of the PDA-based advisor indicates that users form a stronger therapeutic alliance with the animated advisor and rate the health information it delivers as more credible, compared to an equivalent text-only interface or one in which only a static image of the advisor is used.3
The integrated system will be evaluated in a randomized within-subjects design study of free living sedentary adults, comparing real-time feedback with equivalent retrospective (end of day) feedback.
Acknowledgments
This work was supported by grant R21 LM008553 from the NIH National Library of Medicine
References
- 1.DHHS. Physical activity and health: a report of the Surgeon General. Atlanta, GA: 1996. [Google Scholar]
- 2.Bickmore T, Gruber A, Picard R. Establishing the computer-patient working alliance in automated health behavior change interventions. Patient Education & Counseling. 2005;59(1):21–30. doi: 10.1016/j.pec.2004.09.008. [DOI] [PubMed] [Google Scholar]
- 3.Bickmore T, Mauer D. Modalities for Building Relationships with Handheld Computer Agents. Paper presented at: ACM SIGCHI Conference on Human Factors in Computing Systems (CHI); Montreal. 2006. [Google Scholar]