Game Development Reference
In-Depth Information
An Affective Student Model
for Intelligent Gaming
Bickmore, 2001; Bates, Loyall, & Reilly, 1992).
EPAs can be implemented as self-determined
entities, which react according to the changes
in their virtual environment or can be controlled
by a central entity, such as computer-controlled
characters in video games (Mateas, 1997). To attain
a reactive behavior, some EPAs are implemented
using a cognitive theory of emotion as a basis
(Bates et al., 1992), such as the OCC model, or
a personality theory (Conati & Maclaren, 2009),
such as Big Five. Computer and educational games
are comprised of audio and visual media that
serve diverse communication functions, such as
creating a sense of immersion, setting a mood and
decreasing the player's learning curve (Collins,
2008; Malone, 1981). In addition, colors can con-
vey different emotional states and moods (Kaya,
Epps, & Hall, 2004). Research has intelligently
employed color to report the learner's emotional
state in virtual learning environments (Razek,
Chaffar, Frasson, & Ochs, 2006) and classrooms
(Alsmeyer, Luckin, & Good, 2008). Colors have
also been used to express and emphasize emo-
tion (Nijdam, 2005). Our research will focus on
selecting the most suitable feedback responses to
maximize learning or understanding. Therefore,
we choose to modulate game-elements, such as
visuals, audio and colors, to influence the learner's
emotional state and convey cognitive, motivational
and emotional responses. The implementation of
these features is described in the, ' PlayPhysics
design', section.
The theory of emotion, acting as a basis for build-
ing the affective student model, is described in this
section. This model infers the learner's emotions
from cognitive and motivational variables through
observable behavior. The learner's attitudes, be-
liefs and social standards are determined through
answers to questions posed as part of a game
dialogue. In addition, the advantages of applying
Probabilistic Relational Models (PRMs) to student
modeling are outlined.
The Control-Value Theory of
Achievement Emotions
The control-value theory of 'achievement emo-
tions' is a cognitive and motivational theory of
emotion (Pekrun, 2006; Pekrun et al., 2007). It
comprises motivational, cognitive, affective, ex-
pressive and physiological aspects. 'Achievement
emotions' are emotions that arise when performing
academic or achievement activities in which the
attainment of a desired outcome is expected. The
theory signals that control and value appraisals
are the most relevant to determining an emotion.
Control , e.g. self-efficacy, is defined as the per-
ceived control over the activity and the outcome.
Value is related to the perceived importance of the
outcome and the desirability and relevance of the
activity. Appraisals of control that are relevant to
determine 'achievement emotions' are situation-
outcome, activity-control and action-outcome
expectancies. Situation-outcome expectancies
relate to evaluation as to whether an action is or
is not required to attain the desired outcome in the
specific situation. Activity-control expectancies
are related to confidence beliefs when starting
and performing a specific action. Action-outcome
expectancies are related to beliefs that specific ac-
tions can facilitate or can prevent the attainment of
specific outcomes. These outcomes can be positive
or negative. The causal attributions of outcomes
This section describes the research methodology
applied to create and evaluate an affective student
model, which infers emotion from motivational
and cognitive variables using observable behavior.
In addition, the design of PlayPhysics is discussed
and described signaling how game-features will
be modulated to deliver pedagogical responses.
Search Nedrilad ::

Custom Search