FORUM ON EDUCATION
Fall 2001

Physlets: A Tool for Physics Education Research

Wolfgang Christian, Mario Belloni and Melissa Dancy

The impact of instructional software on mainstream physics instruction has, at present, been minimal. At American Association of Physics Teachers (AAPT) meetings in the 1980s, it was common to see participants sharing floppy disks and trading software for the computer-enabled educational reform that everyone knew was sure to come. It didn't arrive, at least not in the form envisioned by the conference participants. Little of the early educational software was adopted by the mainstream teaching community and almost none of it is still being used today. In contrast, printed material from the much earlier post-Sputnik curricular reform movement - the Berkeley Physics series, for instance - is still available and useful to physics educators, although the pedagogy upon which it was based has gone out of fashion. Will this scenario be repeated and are we doomed like the Greek hero Sisyphus to forever push computational physics up the hill of curriculum reform? Can we expect widespread adoption of computation in the current curricular reform initiative? And, if so, what strategies should we adopt to insure that computational-rich curricula being developed today will be adopted and be in widespread use a decade from now? Our approach has been to develop curricular material that couples a software design philosophy with physics education research (PER). It is based on open Internet standards such as Java, JavaScript, and HTML as well as research into the effectiveness of computer-based physics instruction.

Physlets - "Physics applets" - are small, flexible Java applets that can be used in a wide variety of applications [Christian 2001]. Because of their dynamic interactivity, Physlets are ideally suited for interactive engagement methods [Hake 1998, Sokoloff 1997, Thacker 1994] such as Just-in-Time Teaching [Novak 1999] Peer Instruction [Mazur 1997] and Tutorials [McDermott 1998]. In addition, Physlets can also be used as traditional lecture demonstrations and can be given as end-of-chapter homework.

We have developed close to one thousand Physlet-based problems over the past four years in support of a number of introductory physics texts. A selection of these problems is available on the CD that accompanies the Physlets book. More importantly, the Physlets upon which these problems are based are freely distributable for non-commercial educational purposes and are now being adapted to support various curriculum reform initiatives.

Physlets and PER

Physics Education Research, PER, informs us that technology does not necessarily lead to improved learning and that we are just beginning to understand how it is best used. For technology to have a long lasting impact on science education, it will need to be based more on successful pedagogy than on the latest software and hardware. For example, streaming video is currently a hot technology, and both traditional broadcasters and software companies are competing to establish themselves in this market. However, research has shown that merely watching video has little effect on student learning, and it is unlikely that streaming video will change this result. Small cognitive effects have been shown to occur using video clips if the showing of the clip is accompanied with in-class discussion or if the clip is used for data taking and data analysis [Beichner 1997]. Two PER researchers, Aaron Titus [Titus 1998] and Melissa Dancy [Dancy 2001], have used Physlets to study the effect of animation on student assessment and student problem solving ability.

n>

Figure 1: A media-focused projectile problem.

Titus measured student attitudes and problem-solving approaches while they were solving Physlet-based problems. The study distinguishes between media-enhanced problems where multimedia is used to present what is described in the text, and media-focused problems, where the student must use multimedia elements in the course of solving the problem. Titus found that media-focused problems are fundamentally different from traditional physics problems, and Physlets are ideally suited for these types of problems. Consider an example from kinematics. A traditional projectile problem states the initial speed and launch angle and asks the student to find the speed at some point in the trajectory. This problem can be media-enhanced by embedding an animation in the text, but this adds little to the value of the problem. Alternatively, this same problem could be written as a media-focused Physlet problem as shown in Fig. 1 where the student is asked to find the minimum speed along the trajectory. In this case, no numbers are given in the text.The student must observe the motion, apply appropriate physics concepts, and make measurements of the parameters he or she deems important within the Physlet. (A mouse-down enables the student to read coordinates.) Only then can the student "solve the problem." Such an approach is remarkably different from typical novice strategies where students attempt to mathematically analyze a problem before qualitatively describing it (an approach often called "plug-and-chug" and characterized by a lack of conceptual thought during the problem-solving process).

Figure 2: A text-based Force Concept Inventory question.

Dancy used Physlets to probe students' conceptual understanding by using a standard diagnostic instrument, the Force Concept Inventory [Hestenes 1992], in which all thirty static pictures (see Figure 2) were replaced by Physlet-based animations (see Figure 3) [Dancy 2001]. Both quantitative and qualitative data was collected from hundreds of students using the Physlet-based version and the results were statistically analyzed. The study showed that Physlet-based problems are less likely to elicit memorized responses because they allow students to respond to what they see, rather than what they read. Physlets tap into students' intuition and deeply-held misconceptions by eliminating the additional step of translating from words or graphs. In general, students had a better understanding of the intent of the questions when viewing an animation and gave an answer that was more reflective of their actual understanding. Dancy speculates that this may be because the animation looks more like real life than something from a physics textbook.

Figure 3: A Physlet-based Force Concept Inventory question.

Both the Titus and the Dancy studies indicate that while computer-based animation can be used for cosmetic and motivational purposes, they are most effective under the following conditions:

• The animation is integral to the question.
• The student must interact with the animation to obtain data.

The effectiveness of Physlets likely depends on many factors such as how well the task targets known student difficulties, how students use visual cues given by the Physlet, how important visualization is to the given task, and the appropriateness of the Physlet to the given task. Nevertheless both studies show that conceptual understanding is key to solving Physlet problems. Without strong conceptual understanding, students are prone to guess, search for the "right" equation, and lack direction.

Just-in-Time Teaching

Although the media-rich content and interactivity provided by technology such as Physlets can be pedagogically useful, it can lack the human dimension that is important to effective teaching. Computer Assisted Instruction (CAI) has already been tried on very elaborate proprietary systems. It is unlikely to be improved significantly by being ported to the Internet. To be truly effective, the communication capabilities of the computer must be used to create a feedback loop between instructor and student. A new and particularly promising approach known as Just-in-Time Teaching, JiTT, has been pioneered at Indiana University and the United States Air Force Academy and further developed at Davidson College [Novak 1999]. It employs a fusion of high-tech and low-tech elements. On the high-tech side, it uses the World Wide Web to deliver multimedia curricular materials and manage electronic communications between faculty and students. On the low-tech side, the approach requires a classroom environment that emphasizes personal teacher-student interactions. These disparate elements are combined in several ways, and the interplay produces an educational setting that students find engaging and instructive. The underlying method creates a synergy between the Web and the classroom to increase interactivity and allow rapid response to students' problems.

The JiTT pedagogy exploits an interaction between web-based study and an active-learner classroom. Essentially, students respond electronically to carefully constructed web-based assignments, and the instructor reads the student submissions "just-in-time" to adjust the lesson content and activities to suit the students' needs. Thus, the heart of JiTT is the "feedback loop" formed by the students' outside-of-class preparation which fundamentally affects what happens during the subsequent in-class time.

Although JiTT can be implemented fully using technically simple Web-based assignments, incorporating Physlet-based questions can heighten the extent to which student under-standing can be probed and encouraged. The JiTT strategy as applied in physics education is richer for the incorporation of Physlets. Consider, for example, the puzzles shown in Figures 4 and 5. These Puzzles involve nearly the same physics, but what is required of the student in order to solve each puzzle is quite different. In each case, the student must understand the concepts of moment of inertia, torque, angular acceleration, angular velocity, and the relationships between those quantities. In each case, it also behooves the student to draw "extended" free body diagrams to consider the forces and torques involved. The dynamic Puzzle, however, requires some visual analysis and understanding of how the speed with which the mass falls is related to the physics quantities such as angular momentum and moment of inertia. Students are expected to analyze each situation, apply the relevant physics, and answer specific questions. The faculty member then prepares a lecture in response to the student submissions.

Conclusion

Based on our results, we believe that Physlets can be a valuable tool for creating interactive curricular material designed around the needs of the student. We have investigated using Physlets to alter existing curricular material. However, the greatest potential of Physlets will come from using Physlets to ask (and answer) questions in ways which cannot be done on paper [Belloni 2001]. (see also Figure 6.)

Figure 6: Physlet-based JiTT question regarding the quantum mechanical barrier problem. Students are asked to find the potential energy by varying the energy and examining the shape of the free-particle wave function.

The best media-focused problems cannot be correctly solved using "plug-and-chug" methods. The fact that data is not given in the text of the problem requires that students apply proper conceptual understanding to the solution before analyzing data. Therefore, it also seems that Physlet problems may be useful for encouraging a "concept-first" approach to solving problems, where students consider the concepts or principles to be applied to the problem before making calculations. This quality seems to make Physlets well suited for evaluating students' application of conceptual understanding to numerical problems and helping students identify weaknesses in conceptual understanding.

Acknowledgements

Portions of the work presented here are based on published and unpublished work in collaboration with Melissa Dancy, Aaron Titus, and Evelyn Patterson.

The author would like to acknowledge the National Science Foundation, grant DUE-9752365, for its support of Physlets.

References

[Christian 2001] "Physlets: Teaching Physics with Interactive Curriculum Material," W. Christian and M. Belloni, Prentice Hall, Upper Saddle River, (2001) See also: http://webphysics.davidson.edu/applets/applets.html

[Hake 1998] R. R. Hake, "Interactive-engagement vs. traditional methods," Am. J. Phys., 66, 64--74 (1998)

[Sokoloff 1997] D. R. Sokoloff, "Using Interactive Lecture Demonstrations to Create an Active Learning Environment," The Physics Teacher, 35, 340 (1997).

[Thacker 1994] B. Thacker, "Comparing Problem Solving Performance of Physics Students in Inquiry-based and Traditional Introductory Courses," American Journal of Physics, 62, 627-633. (1994).

[Novak 1999] "Just-in-Time Teaching: Blending Active Learning with Web Technology," G. M. Novak, E. T. Patterson, A. D. Gavrin, and W. Christian, Prentice Hall, Upper Saddle River, 1999.

[Mazur 1997] "Peer Instruction: A Users Manual," E. Mazur, Prentice Hall, Upper Saddle River, 1997.

[McDermott 1998] "Tutorial in Introductory Physics," L. McDermott and P. S. Shaffer, Prentice Hall, Upper Saddle River, 1998.

[Beichner 1997] R. Beichner, "The Impact of Video Motion Analysis on Kinematics Graph Interpretation Skills," American Journal of Physics, 64, 1272-1277. (1997).

[Titus 1998] A. Titus, "Integrating Video and Animation with Physics Problem Solving Exercises on the World Wide Web," Ph.D. dissertation. North Carolina State University: Raleigh, NC. (1998).

[Dancy 2001] M. Dancy, "Investigating Animations for Assessment with an Animated Version of the Force Concept Inventory," Ph.D. dissertation. North Carolina State University: Raleigh, NC. (2001).

[Hestenes 1992] D. Hestenes, M. Wells, and G. Swackhamer, "Force Concept Inventory," The Physics Teacher, 30, 141--158 (1992).

[Belloni 2001] M. Belloni, L. Cain, and W. Chrisian, "Using Physlets and Just-in-Time Teaching in Quantum Mechanics," AAPT 2001 Summer Meeting, Rochester, NY, 2001. See also: http//webphysics.davidson.edu/qmbook/

Wolfgang Christian, Mario Belloni and Melissa Dancy are at Davidson College, Davidson NC 28036. W. Christian is a Professor Physics. His research interest is in the area of computational physics and instructional software design. He is co-author of Physlets: Teaching Physics with Interactive Curricular Material (2001), Just-in-Tiome Teaching (2000), and Waves and Optics (1995). M. Belloni is an Assistant Professor of Physics and his research interests are in theoretical physics and physics education. M. Dancy is a Visiting Assistant Professor with research interests in physics education and instructional technology.