Affiliations: Department of Computing, Faculty of Science, Macquarie
University, Sydney, Australia | Northeastern University, Shenyang, Liaoning, China,
110819
Abstract: Exam stress is a common predicament faced by students of all age
groups and cultures. If improperly managed, it can lead to insomnia,
depression, suicide and many other negative health implications for students at
all levels of education. However, exam stress management gained relatively low
attention from the society and academia. Stress in optimal level can stimulate
students to achieve their personal best in their exams. However, excessive
stress can be devastating for their well-being. Consequently, it is highly
imperative that timely counselling services are made available, particularly
during their exam period when there is a bursting demand to access these
resources. In this study, we propose that a virtual intelligent university
student advisor would be highly useful and effective in meeting this demand. We
have developed an embodied conversational agent, ESCAP that simulates the
facial animations and advice of a professional psychologist to support
undergraduate students in stress management during their exams. In this paper,
we discuss the system architecture of ESCAP as well as the practical and
theoretical implications of the system. We conducted a number of pilot tests
with 25 students to design the affective system and found in 200 samples that
in general the voices of male advisors are considered as more pleasant and
credible than female advisors'; on the other hand, the voices of female
advisors are considered as more clear, dynamic and competent than male advisors
by most users. These findings have important implications on the design of
virtual psychologists, especially regarding the gender of the virtual advisor.
System developers should consider the gender of virtual advisors carefully to
maximize their effect on users. The next stage of the project involves large
scale experiments testing ESCAP's impact on students. In these experiments, we
plan to measure the stress level of students using both qualitative and
quantitative measures such as heart beat and skin conductivity before and after
their interactions with ESCAP.
