Teaching Java and Abstract Data Type Semantics with
Multimedia*
William M Pottenger, Assistant Professor, pottenger@cse.lehigh.edu
G. Drew Kessler, Assistant Professor, kessler@cse.lehigh.edu
Edwin J. Kay, Professor, kay@cse.lehigh.edu
Computer Science and Engineering Department
Lehigh University
19 Memorial Drive West, Bethlehem PA 18015
CIMEL is a
multimedia framework for Constructive and collaborative, Inquiry-based
E-Learning supplementing computer science courses. Constructive learning
goes beyond learning by receiving knowledge, to learning by building systems,
with immediate, visual feedback. Collaborative learning encourages
students to interact with instructors and librarians, via live links and
remote-controlled “show me” sessions and by reviewing multimedia FAQs of
recorded “show me” sessions. Inquiry-based learning guides the student
into pursuing exploratory research in a community of students and scholars. A
text mining and visualization tool also enables students to identify and
explore emerging technology trends in computer science as part of our
inquiry-based framework. Our project documents, evaluation materials and a
prototype are available at www.cse.lehigh.edu/~cimel.
Within this
framework, we have developed alpha versions of new materials for two courses
in computer science at quite different levels: a graduate level course in
Object-Oriented Software Engineering (OOSE) and a first semester course in
computer science (CS0/1). New multimedia
modules, implemented in Flash and playing through high-speed connections on
the web, feature audio narration, animation, simulations, quizzes, and constructive
exercises. Two different experiments
have been developed to evaluate the new materials.
For the OOSE course, we developed a multimedia unit on Abstract Data Types (ADTs), as a way to formalize the meaning of classes in connection with
object-oriented design. In the past, graduate students have found it difficult to master this material from lecture and textbooks alone. We have designed an experiment
to determine whether multimedia actually improves learning, both in terms of objective knowledge and ability to perform a task, designing ADTs for a sample problem
involving several classes and inheritance. We divided the 20 students into two groups, A and B. Both groups took the same pretest of 20 questions, with questions
and choices presented in random order. Then group A studied the multimedia. Then both groups took a mid-test, consisting of 20 questions, each covering similar
topics as the original set, again with questions and choices presented in random order. (Both groups took the mid-test so that the practice effect could be accounted
for in our analysis.) Group A performed significantly better, with mean scores improving from 9.2 to 15.6, while group B showed no change. The difference between
groups was highly significant, F(1,20)=42.23, p<0.0001. Both groups then heard a lecture on ADTs and assigned reading from relevant chapters in a textbook
on reserve at the library. Then both groups were given an assignment, creating ADTs to solve a problem involving several classes, inheritance and dynamic binding.
After handing in their assignments, Group B saw the multimedia. Then both groups were given the opportunity to resubmit the assignment. Our hypothesis was
that Group B would be more likely to resubmit the assignment with improvements. Several students did so, but the overall results were not statistically significant,
F(1,40)=3.25, p>.05. Both groups then took a post-test of another 20 questions. Group B improved significantly, while Group A’s performance remained the
same as the mid-test, as measured by a test for the group-x-test interaction, F(2,60)=7.77, p<.001. This result suggests that the multimedia contributes to objective
learning regardless of whether it is presented before or after the lecture, and could conceivably be used independent of lecture. The results for task learning are
less clear, but the study has provided useful input about improvements to the multimedia; more guidance will come from a web-based survey evaluating the alpha
version and a focus group.
For CS0/1 (CS0
for non-majors and CS1 for potential minors and majors), we have revised both
the manuscript and multimedia of The Universal Machine: A Multimedia Introduction
to Computing (McGraw-Hill 1998), tentatively retitled The Universal
Computer: A Multimedia Introduction to Computer Science and Problem Solving
in Java. As the new title suggests, the new material introduces Java (instead
of C++). This fall, students in the
introductory course at Lehigh University began using the new material, with
five of the six proposed chapters introducing Java programming and all ten
other chapters on problem solving, software engineering and the breadth of
computer science. To evaluate the
new material, we divided students into two groups: an experimental group (80%
of about 50 students consenting to participate) getting access to all materials
(the complete new manuscripts, the old multimedia associated with The Universal
Machine, and new multimedia for five chapters on Java programming associated
with The Universal Computer), while a randomly selected control group
(20%) gets everything but the new multimedia. We are comparing performance of the two groups,
with respect to performance on actual Java programming assignments and examinations,
as well as a pretest (administered at the beginning of the course) and a posttest
(near the end), consisting of the same 30 multiple-choice questions, with
questions and choices presented in random order for each participant.
A survey solicits feedback from the students with respect to the user
interface and content of the new multimedia.
Demographic data has been gathered in order to track performance and
feedback of different subgroups, such as women and minorities.
We will present the results of this study at the SIGCSE conference.
Preliminary
versions of our collaborative learning and emerging trends detection tools
have also been developed and studies evaluating their effectiveness are under
way. We plan to share results of these
studies at other conferences.
We plan to conduct another experiment evaluating a beta
version of the Abstract Data Type module next semester, in a class of sophomores
and juniors. We also plan to develop new modules for both
the OOSE and CS0/CS1 courses. Beta
versions of the multimedia, as well as a revised draft of The Universal
Computer, will be available for use in courses at other educational institutions,
in the fall of 2002. Any parties interested
in evaluating our materials for possible adoption should contact the lead
author.
*
This project is funded by National Science Foundation (Grant No : EIA-0087977)