Graphics Reference
In-Depth Information
the intervention: on the one hand, it addressed the difficulty of the
actual situation and the emotional state of the user and, on the other
hand, it communicated on a more interpersonal level. In providing the
intervention, the system created a relation offer to the user through
empathy, which he or she could accept or deny. We can therefore
assume that a user will only accept the system's relation offer if it is in
keeping with his/her of mental states to the system. An intervention
which is inappropriate to the situation or the relation the user has
with the system will be unacceptable.
The quantitative analysis of speech behavior showed a tendency
of the users' to ascribe human characteristics towards the system. In
order to respond individually to users, more information
is needed
about users' individual experiences when ascribing. The analyses of
the interviews gave insights into the ranges and variances of these
experiences. It became apparent that these results sometimes differed
from the quantitative analysis.
The ascriptions identified during our interview analysis, as well
as their effect on users' behavior, have led to numerous resulting
theoretical considerations according to their relevance: On the one
hand, the ascriptions can increase users' co-operation and trust in
the system, if they ascribe positive personality traits, motives, aims,
attitudes, etc. (for example willingness to personally identify the user).
On the other hand, negative ascriptions can lead to a reduction of co-
operation, accompanied by feelings of mistrust and skepticism (like
ascribing the possibility of abuse of confidence to the system), which
could lead to reactive user behavior.
All in all, our results seem to fortify our assumption, that beyond the
programmer, who implements personality within computers, it is the
user who is “giving” personality to the computer by ascribing different
mental states like motives, aims, feelings, as well as personality traits
to it. Further investigations are needed, especially regarding relations
between the linguistic features of users' utterances, as well as users'
experiences when ascribing during HCI. Moreover, investigating the
impact of different system interventions on these relations, which
should be individualized for users or user-types, seems worthwhile
for maintaining and increasing users' co-operation in HCI.
Acknowledgements
The presented study is performed in the framework of the Transregional
Collaborative Research Centre SFB/TRR 62 “A Companion-Technology
for Cognitive Technical Systems” funded by the German Research
Search WWH ::
Custom Search