Graphics Reference
In-Depth Information
physiological agitation, and facial movements through sensors in the
chair, the monitor, the mouse, and the skin. A relatively simple model
is used, in which the values of four parameters can be allocated to
the following four dispositions (here referred to as emotions) with
a precision ranging from 78% to 87.5%: boredom, flow, interest, and
frustration. Empirical tests have shown that by using emotionally
adequate responses (50 variations of support and encouragement, to
keep going and trying hard), students work significantly longer on
frustrating and difficult tasks and that their stress level is decreased.
The tutorial systems not only respond verbally, but are also able to
show an emotional response to the users in the form of avatars, or
support the interaction with an interested facial expression as well as
positive gestures. The empirical research has shown significant gender
differences: Female test subjects benefited more from the emotion-
based responses of the digital tutor than male test subjects.
Simply because of the fact that language is the most important
form of communication for human beings shows that companion
systems should also be able to communicate verbally. Nass and
Brave (2005), however, impressively documented which possible
consequences must be considered when technical systems are
equipped with verbal interaction capabilities. Verbal technical systems
could potentially be associated with certain social competencies or
personality traits that could have a conscious or unconscious influence
on the user's interaction with a technical system. If the expectations
are not fulfilled, the system might be less accepted and not trusted.
Notwithstanding these risks, however, verbal communication also
offers the unique opportunity to increase trust and acceptance not
only by providing information in an effective and natural manner,
but also by using voice modulation (prosody) capabilities. Humans
are probably more aware of what is being said in a conversation,
but how something is said is often just as important, especially in
social interactions. In the tutorial area, the prosody tool is used to
positively influence the learning situation. Prosody can, for example,
support the motivating character of comments and help in learning
situations to increase perseverance and avoid errors. Consequently,
this tool also seems suitable for interactions with technical systems
to increase effectiveness. While there have been numerous studies in
recent decades about the production and processing of prosody in
human-human interactions (Frick, 1985; Scherer, 2003b; Baum and
Pell, 1999; Friederici and Alter, 2004), this topic has not yet been
investigated intensively with regard to the human-machine interaction
area. Until now, the recognition and classification of human
prosody by the technical system have been the main topic of the
Search WWH ::
Custom Search