Department of Computer Science
Vanderbilt University, Nashville, Tennessee 37235
e-mail: thaddeus@vuse.vanderbilt.edu
The computer-based learning environments being developed in this project
has been named Macrocontext Plus Microworlds (MPM),
indicating that traditional microworld exploration is anchored
within a complex, realistic problem solving situation, providing the student
with a comprehensive focus and objective in developing a solution and
comparing alternative solutions.
A Jasper MPM for trip planning problems, called AdventurePlayer,
has been under development for the last three years.
The system allows students to construct, evaluate, and
reason about solutions to trip planning problems.
AdventurePlayer incorporates a number of tools
to facilitate the complex trip planning problem solving task,
including an icon pallet, a planning notebook, a timeline,
a plan simulator, and a planning coach (see FIGURE).
One of the goals of the AdventurePlayer ILE
is to use representations that make explicit concepts such as
sequencing dependent actions and performing independent actions
in parallel to reduce the time required to achieve the goal.
In addition, the system's ability to
generate real-time analogous problems affords the student
opportunities to examine the relationships between
problem variables.
Instead, AdventurePlayer is a learning environment designed to
cultivate the intelligence of active, intentional learners [5].
The AdventurePlayer environment is situated within the epistemological
and didactic framework of situated cognition (cf., [3]).
Students perform problem solving in the context of a meaningful anchoring
situation, thus providing focus and direction for their discovery
learning. At the same time, the student may have access to a
number of additional microworlds relevant to the complex problem.
One could conceive of additonal microworlds that
enhance scientific and a historical understanding, such
as a physics microworld for exploring the relationship between thrust,
payload, and lift caused by the shape of the wing, an engineering
microworld that studies combustion engines, and historical microworlds
that revisit significant events in aviation, such as Lindberg's solo
Spirit of St. Louis flight across the Atlantic.
While MPMs share some exploratory theory with increasingly complex
microworlds (ICM) [4], some distinctions are apparent:
In the timeline/simulation group,
77% of the participants were able to derive a complete solution.
In the group using the unembellished system,
only 8% generated complete plans.
Of the 17 Group 2 students generating complete solutions, 5 arrived
at the optimal solution without any coaching. The remaining 12 received
coaching, of which 58% achieved the optimal solution with only one
coaching interaction and 17% achieved the optimal solution after two
coaching interactions. The remaining 25% did not arrive at the optimal
solution.
The analysis discussed above suggests that there was indeed value-added
by the timeline, simulation, and coaching capabilities of the system.
Furthermore, Group 2 students appear to have performed better on a posttest
transfer task involving a multistep word problem outside the trip planning
context.
Abstract
Intelligent Learning Environments (ILE) are constructivist systems that attempt
to incorporate beneficial aspects of tutoring systems and cognitive
tools. ILEs support discovery learning through reflective
interaction as well as curriculum-driven learning through scaffolding
and coaching.
ILEs are concerned with students developing both general
and domain specific thinking and problem solving skills.
AdventurePlayer is an ILE designed
to facilitate constructivist learning in the context of an
anchored instruction curriculum.
Keywords:
intelligent learning environments, anchored instruction,
macrocontext microworlds, trip planning, optimal solutions,
heuristic techniques.
Introduction
The Cognition and Technology Group at Vanderbilt (CTGV) is developing
The Adventures of Jasper Woodbury Series , a sequence of video-based
adventures designed to motivate students and
help them learn to think and reason about complex, real life problems.
The theoretical framework for the project is consistent with constructivist
theories and emphasizes generative learning anchored in meaningful contexts.
These contexts are presented in the form of a story on videodisc,
and are called "macrocontexts"
because they involve complex situations that require students to formulate
and solve a set of interconnected subproblems [3].
By comparison, problems found at the end of textbook chapters often
involve a series of disconnected "microcontexts." The CTGV
anchored instruction approach resembles case-based and
problem-based programs
and affords some of the advantages of "in-context" apprenticeship
training [2].
DESIGN ISSUES
AdventurePlayer is an exploratory environment designed to
assist learners in the activity of constructing knowledge in a manner
that augments--not replaces--the role of the teacher. The system
is not a traditional Intelligent Tutoring System with full system
control and detailed student modeling based on sensitive cognitive
diagnoses [1].
SYSTEM EVALUATION
An evaluation study of the AdventurePlayer system was recently
conducted. A total of 48 sixth grade students participated in one of
two conditions. Group 1 worked in a basic AdventurePlayer MPM that
included three of the capabilities of the system: (a) the icon pallet
for accessing information, (b) the planning notebook for generating and
editing a plan, and (c) the map from the video that could be used for
referential information. Group 2's MPM further included the timeline
and simulation capabilities. As well, if Group 2 students developed
complete but non-optimal plans they received coaching designed to
foster a more optimal plan. All students initially viewed the RBM
adventure and then worked in
the appropriate MPM to solve it. The design thus permits us to test the
value-added of the timeline, simulation, and coaching capabilities.
CONCLUSIONS
The AdventurePlayer system is an Intelligent Learning Environment
situated within the epistemological and didactic framework of situated
cognition. Problem solving is supported by a number of components
including a planning notebook, timeline, simulator, and coach.
Evidence to date indicates that AdventurePlayer will provide
beneficial results to performance in a domain known to be problematic.