The Physics Illumination
Project: Conceptual Homework on the Web
Physics Illuminations consist
of a simple interactive component (such as a Java applet) packaged
with brief descriptive text. Most are qualitative in nature and naturally
incorporate student assessments, so that they work well as automatically
graded conceptual homework. Their positive impact upon learning has
been demonstrated by the fact that over 100 students in three different
classes employing no in-class interactive engagement methods showed
conceptual learning comparable to that of the interactive engagement
classes reported by Hake.1-2 (Further tests with other
instructors are currently underway with the support.)
In-class active learning methods
have been widely studied and are generally believed by physics education
reformers to substantially improve student understanding of basic
physics concepts. However, because of the extensive attention that
has been paid to in-class methods, it is unlikely that substantial
further improvement can be made on that front (other than convincing
more instructors to use an interactive approach). By contrast, comparatively
little attention has been paid to what can be done with homework
to improve student conceptual learning. For typical students, learning
through out-of-class study is probably only weakly correlated with
learning through interactive in-class techniques; consequently, the
combination of effective homework with interactive engagement methods
may help us further improve student learning in introductory physics.
Listed below are a number of
my conclusions about what is needed for effective web-based homework,
based on my experience with Physics Illuminations. (Some of these
items have been previously discovered by others in the context of
in-class learning studies.)
- Most students will use on-line
learning materials only if such use directly affects their grade.
Giving them evidence that students who voluntarily use the materials
score higher on quizzes or exams is not a sufficient inducement.
- An applet should be accompanied
by text to accommodate less explorative learning styles. However,
such text should be very brief for students to take the time to
- The opportunity to score
well on homework encourages a high rate of participation, and consequently
more learning. This is one of several arguments for relatively
simple, single-focus learning items.
- Applets that focus on a single
topic and limit the number of variable parameters are more likely
to be effective. Most students are overwhelmed by an applet that
gives them control over many variables, since they have not internalized
the scientific approach of studying the effect of one variable
at a time.
- Approaching a concept from
different directions (with different applets) helps solidify the
learning of that concept.
- Students can learn effectively
through repetition. Applets that present students an "unlimited" number
of random cases are particularly suitable for computer-aided learning.
- It is not necessary to show
students what they did wrong if the software can give multiple
random cases of simple tasks. Students will read the accompanying
text (if it is brief) to find out what they need to do or what
they might be doing wrong. Of course, immediate feedback is essential,
but that is a given in computer-assisted learning.
- For complicated tasks, such
as problem solving, and tasks where few cases are available, context-sensitive
feedback is likely to be necessary. This is a much more difficult
research and programming problem, requiring substantial understanding
of learning (by both human and machine). Thus, it is important
to break such tasks into simple subtasks to whatever extent is
Improving student learning of
physics through the use of more effective homework is still a relatively
unexplored area. If you are interested in getting involved in this
exploration, check out the open source Physics Illumination Project
I welcome participation by non-programmers as well as programmers.
Partial support for this
project is provided by the National Science Foundation's Course,
Curriculum, and Laboratory Improvement Program under grant DUE-0088695.
- Ronald L. Greene, "lluminating
Physics via Web-Based Self-Study," Phys. Teach. 39,
356-360 (Sep. 2001).
- Richard R. Hake, "Interactive
Engagement versus Traditional Methods: A Six-Thousand-Student Survey
of Mechanics Test Data for Introductory Physics Courses," Am.
J. Phys. 66, 64-74 (1998).
Ron Greene is a Professor
of Physics at the University of New Orleans. During his career
he has had varied research interests, among them plasma spectroscopy,
semiconductor physics, and machine learning. This last area has
evolved into studies in computer-assisted instruction.