Game Development Reference
In-Depth Information
Figure 9. Comparison between predicted emotion and reported emotion
tory joy', the students obtained a mark correspond-
ing to a high level of performance. However, the
students may report 'anxiety' due to the way that
the evaluation process was carried out, since they
were located in the same room at the same time,
they may have thought that their performance
would be available to everyone. Therefore, in
successive evaluations, the students will be
evaluated in isolation in the Gesell dome.
hancing ITSs and intelligent game-based learn-
ing environments through building an affective
student model. This model infers emotion from
cognitive and motivational variables using the
learner's observable behavior and employing the
control-value theory of achievement emotions. A
Probabilistic Relational Models (PRMs) approach
to building the innovative affective student model
was discussed. PRMs facilitate the design process
of Bayesian Networks (BNs), since they reduce
the required complexity. A preliminary evalua-
tion was carried out with seven participants to
identify deficiencies in the prototyping material.
An accuracy of 57.14% was achieved when the
reported-emotion by the students was compared
with the predicted emotion by the system. These
results may be influenced by the probabilities set
in the Conditional Probability Tables (CPTs). To
model accurately the CPTs and evaluate effectively
the model a larger population of participants
would be necessary. Students signaled some
recommendations to enhance the effectiveness
of the prototyping material, which will be taken
into account when performing future evaluations.
Future work includes testing the accuracy of
the affective student model through a prototyping
exercise involving a large number of undergradu-
ate students enrolled in an introductory Physics
course. The objective is to determine if the iden-
tified variables can ensure the accurate inferring
CONCLUSION AND FUTURE
RESEARCH DIRECTIONS
Intelligent game-based learning environments or
educational games are inherently effective moti-
vational tools. However, attaining the learner's
attention and engagement is simply not enough.
Educational games have yet to prove effective
at teaching and must incorporate assessment
mechanisms and suitable feedback to support
independent learning. Incorporating Intelligent
Tutoring Systems (ITSs) into educational games
is a plausible solution to these problems, since
ITSs can keep track of each student's history of
interaction and adapt accordingly to the identified
student needs. The advantages of Affective Gam-
ing, such as evaluating the player's experience
and understanding when an emotion is relevant,
were emphasized. Our research focuses on en-
Search Nedrilad ::




Custom Search