usually totaled
and averaged to produce a grade for the test. At the end was a grade,
which did not say much about what the student really learned.
Using the expert system shell, I create a multiple- choice question
that branches to different questions based on the students answer.
A single multiple-choice question becomes a nucleus of knowledge that
assesses the student's knowledge as well as what further knowledge
should be tested.
For example, look at this multiple-choice question.
The central processing unit (CPU) in a Z-80 microcomputer
contains
The
a) ALU
b) ROM
c) RAM
d) I/O ports
If the student answers "a) the ALU", then we have
assessed that the student understands at least the architecture of
the Z-80 chip containing the ALU. We also assessed that the student
knows that the ALU is physically separate from ROM, RAM and IO ports.
Once saved, possibly in a database, the
information associated with the response "a) the ALU" determines
the next question to present. Not just any question should follow;
only a question that assumes the student has the knowledge represented
in the current question. The system moves forward to the next question,
based on knowledge acquired from the student.
An interesting thing happens if the student
answers "b) ROM". The expert system shell could optionally
present just-in-time lessons, first on ROM and then on the Z-80 ALU.
When the student answered "b) ROM", we not only determined
that this was a wrong answer, which an ordinary multiple-choice test
would do, but we also determined that the student didn't understand
the Z-80 ALU or the ROM nor the relationship between the two. Therefore,
we could branch to an associated ROM question, which may lead to an
appropriate ROM related lesson. If the student answered "c) RAM",
we could branch to a question on RAM, which may lead to the presentation
of an appropriate lesson related to RAM. This is possible for every
distracter.
What the test developer has created here goes
well beyond a multiple-choice test. It becomes an assessment driven
adaptive tutorial system that continually adjusts precisely to the
individual student's knowledge. When the user answers the questions
|
|
Intelligent Machine
Inc.
O'INCA
Design Framework for Windows
Integrated environment for development
of intelligent
adaptive systems.
|
FUZZY LOGIC
|
EASY TO USE
GUI &
DESIGN DOCUMENTATION
|
|
NEURAL NETWORK
|
SIMULATION
& DEBUGGING
|
|
USER-DEFINED
|
VALIDATION
& CODE
GENERATION
|
|
|
DECISION SUPPORT &
REASONING SYSTEMS
PROCESS CONTROL, PATTERN RECOGNITION
SYSTEM MODELING
|
|
$1,895 Enterprise
Version
|
$1,295 Educational
Version
|
|
Intelligent Machine,
Inc.
www.OINCA.net
Email info@oinca.net
Tel (408) 230-6441
|
|
|
| correctly, the
system moves on to the next question. When answered incorrectly, the
system can present appropriate training material to correct the misconception
and present appropriate questions regarding the specifics of the incorrect
answer. If the student continues to answer questions incorrectly,
the system can present additional tutorials and questions until the
student reaches the most elementary presentation materials. Conversely,
as the student continues to answer questions correctly, they advance
through the test in an efficient manner, perhaps never receiving just-in-time
tutorial material, and perhaps skipping many questions. |
