Physics Education Research Group

Progress Report

Digital Video Interactive - a Case Study in Physics

NSF Grant Number MDR 9150222

Submitted by Dean Zollman, Kansas State University

February, 1994

This project has continued to explore the use of digital video in the teaching of physics and to develop materials and tools for using digital video. Over the last two years, digital video has become much more available in science classes. This availability is primarily the result of a significant increase in interest in multimedia computing and a corresponding decrease in the cost of digital video equipment. Because of the relatively low cost video capture and playback boards which are now available, many more teachers and students are interested in digital video as a laboratory tool than were interested in it at the beginning of this project. Thus, we have emphasized video as a laboratory tool (or video based laboratories) as a primary use for digital video.

This emphasis connects well with our previous work in which we use interactive videodisc in laboratory setting. A review of the use of interactive video in physics teaching and of video in general in a laboratory setting is contained in a paper by Dean Zollman and Robert Fuller. This paper, which summarizes some aspects of the present project as well as several other projects which have been carried out over the last 15 years, is tentatively scheduled for publication in the April issue of Physics Today. A copy of the paper is attached as Appendix A to this report.

Most of the efforts in this project center on using video as a way to visualize ideas in modern physics. Our major effort has concentrated on relativity and quantum mechanics. The relativity materials concentrate strictly on using video as a laboratory tool and having students collect data and make observations via video. These materials have been tested on a variety of students ranging from fifth and sixth graders to elementary education undergraduate majors to graduate students in educational technology. The quantum mechanics materials are not as complete at this moment as the relativity materials. To date they have been tested primarily with students who are not majoring in physics and are taking a course in contemporary physics as an elective. The types of students in this course range from nuclear engineering majors through elementary education majors. Thus, our materials have been tested with a broad range of students most of whom are not majoring in a science or engineering.

Video Based Laboratories (VBL) and Video Processing

Much of our work with video processing and its use in teaching reference frames is summarized in a paper by S. Raj Chaudhury and Dean Zollman which has been submitted to Computers In Physics. That paper is included as Appendix B to this report.

Since completion of that paper, we have developed two Windows-based, interactive programs for the capture, playback, and analysis of digital video, and have incorporated both of them into activities in which students gain experience in applying the concept of frames of reference. The first of our programs uses Intel's ActionMedia II package to capture video to the hard drive of a computer. Its playback options include continuous playback, frame increment, or random frame selection using a ``slider'' control. The slider allows video update far faster than anything else we have seen, providing virtually instantaneous frame advance when operating on a 486 PC. The mouse can be used to position the cursor anywhere within the frame, leaving a mark at that location and writing the corresponding image coordinates to a file in a format readable by most spreadsheet programs. This allows users to delineate the path of a moving object on the video screen, and to perform calculations and graphical analyses of these paths using the generated coordinate file. Modifications currently being made to this program include varying the size of the video window, providing more flexible control of the video playback, and more complex labelling of points and lines of interest.

The second of our programs does not capture video, but serves as a tool for analyzing linear motion in detail. Also using the ActionMedia II file format, this program allows the user to define a rectangular region with the mouse, in effect cropping each frame to retain only the region inside of which the motion of interest takes place. The cursor can then be used to designate two objects inside the specified rectangle as objects to be located by the computer in all subsequent frames. When the two objects have been digitized, the program will extract the same region from each frame and place them adjacent to one another either beside or beneath the window in which the video is simultaneously being displayed. Two white points appear in each rectangle to indicate the location of each of the two objects of interest at that time. The trails of white dots that result from the accumulation of these rectangles represent space-time diagrams plotting distance versus time. After constructing a space-time diagram for the laboratory reference frame, the user can view the space-time diagram corresponding to the same interaction observed from the frame of reference of either of the two digitized objects, the center of mass reference frame, or a frame corresponding to a user-defined Galilean transformation. Modifications are being made to this program to include the ability to follow the motion of an arbitrary number of objects, each of which will be provided with a name by the user and marked on the video screen by a unique color. The transformation of space-time diagrams to that observed from the frame of reference of any of these objects will then be possible. In addition, the rectangular region of interest designated by the user will be of arbitrary orientation, rather than being either horizontal or vertical as it is now. Eventually, we intend to generalize the detection algorithm to accommodate two-dimensional motion within the entire frame.

Around these programs, we have constructed a series of three activities in which students capture video of experiments they perform and then use one or both of the programs to analyze their videos. In so doing, they are able to examine how the appearance of an object's motion depends upon the frame of reference from which it is observed. Each of the three activities, entitled ``Space-Time Diagrams,'' ``The Ball Drop and Frames of Reference,'' and ``The Human Cannonball,'' is accompanied by an activity packet which presents as a storyline a real-life problem which is to be modelled by the activity. The packet also includes instructions for carrying out the activity and using the equipment, and questions to be answered at each stage of the activity. A copy of the activity packet is attaches as Appendix C.

We have tested these activities with pre-service elementary teachers in an introductory-level physics course at Kansas State University taught during the fall semester of 1993. By the time students arrived for these activities, they had been presented with the concept of frames of reference in the course. Each activity was operated over a two-week period during which 24 groups of three students worked together for about an hour, performing the experiment and answering the questions as a group. All three activities were done using an IBM Model P75 486. Preliminary results indicate that students who participated in all three activities significantly enhanced their understanding and ability to apply the concept of frames of reference. Students who participated in one or two of the activities also benefitted greatly from the experience, but we believe that the most beneficial results occur when all three activities are done, in the order presented above.

One of the most interesting tests of the relativity materials was completed in science clubs which meet after school in the elementary schools in Manhattan, Kansas. Elementary education majors who had completed some of the digital video relativity experiments developed similar materials for use with their science club students. They, then, presented this material as a laboratory activity to some of the science clubs while others performed similar activities without the digital video component. Data were collected on students attitudes and knowledge from the activities. Unfortunately, the data analysis is not yet complete. Some aspects of this effort are described in a paper by Dean Zollman which has been submitted to Physics Education at the invitation of the editors. (A copy of that paper is attached as Appendix D.)

Visualizing Quantum Mechanics

Our approach to quantum mechanics has attempted to combine two types of materials. First, computer materials to visualize the wave function and how a wave function or probability distribution looks under certain circumstances dates back to the 1960s. These computer materials have usually been created for upper level undergraduate or graduate physics students. We have modified these approaches to make them more palatable for an introductory student, particularly, a non-science student who may be taking only one course in physics. At the same time, we wish to emphasize that many of the results of quantum mechanics are a natural consequence of deciding that small objects must be treated as waves. Thus, we are using digital video to help students see the analogies between mechanical waves and quantum waves.

To illustrate principles of quantum mechanics we have developed three Windows- based computer programs that solve complicated equations and provide graphical illustrations of their solutions. The package is contained within a presentation developed using Microsoft's Multimedia Viewer. The presentation includes a tutorial on the basics of quantum mechanics illustrated by videos of mechanical analogs and our three programs, all within a hypermedia environment. Students are thereby able, while reading the material, to move smoothly between related topics, videos, figures, and glossary definitions by clicking the mouse on highlighted text and pictures.

The first of our three programs illustrates Heisenberg's Uncertainty Principle by using Fourier methods to construct wave packets from discrete momenta, providing a graphical depiction of the resultant wave packet in both real space and momentum space. The second program solves the time-independent Schrodinger equation for arbitrary potentials. It includes a small set of built-in potentials and allows the user to input potentials either by drawing them with the mouse or by reading values from a file generated by a spreadsheet program. One mode of operation allows the user to input energy values and observe the resulting wave functions. In the other mode, the computer solves for energy eigenvalues and displays the corresponding wave functions. The third program solves the time-dependent Schrodinger equation for arbitrary potentials. It also allows input of potentials either by drawing or by reading from a file and allows the user to vary scaling parameters. A simulation of a scattering event is then animated in a separate window, showing the time evolution of the wave function.

An additional effort to visualize quantum effects involves the creation of models of electron clouds which undergo transitions from one state to another. The creation of these models as simulations on a personal computer is beyond the capabilities of present day computers. Thus, we have used a Silicon Graphics workstation to create the images and then convert them to digital video. These videos can then be transported to the personal computers as QuickTime or Video for Windows files and played back as a model of the atomic processes. With the cooperation of the high-performance computing facilities at Kansas State we will render higher quality versions of the videos by using some Sun SparcStation 10 computers. We will convert again these animations to Video for Windows and QuickTime formats so that they can easily be ported to PC and Macintosh platforms. Investigations of ways to make these videos more interactive are in their initial stages.

Initial testing is being undertaken with students taking a course in contemporary physics at Kansas State. A copy of the written materials related to Visual Quantum Mechanics is attached as Appendix E.

Miniature, Infrared-Sensitive Video Cameras

We have acquired a miniature video camera that has enhanced sensitivity in the near-infrared region of the electromagnetic spectrum. With dimensions of approximately 11cm by 6cm by 3cm, the camera requires a 12-volt power supply and provides a standard video signal as output. Perhaps the simplest illustration of the camera's sensitivity is the bright pulse of light observable on an attached monitor when any common infrared-LED remote control is pointed at the camera and activated. We have done experiments comparing images taken using this camera and an otherwise-normal camcorder, examining burners on a stove top, heating elements in an oven, and a laboratory heating coil with a range of applied voltages. The results show in a very straightforward manner that the infrared camera does detect wavelengths longer than can the camcorder, thereby making it a useful tool for demonstrating to students the existence of infrared radiation and illustrating common infrared sources.

Spectral Analysis Using Video Capture

During the summer, 1993, we were fortunate to have an undergraduate scholarship student who was funded by a grant from the Howard Hughes Medical Foundation. This student had an interest in the spectral response of video cameras and the possibility of using video capture to complete a spectral analysis. We hoped that we could use "off the shelf" components and software to do such an analysis for relatively low cost. Unfortunately, this experiment was only partially successful. Part of our difficulty stems from a lack of knowledge about the true spectral response of the video camera. While it is possible to obtain spectral response for the charged couple device in the camera, we are not able to obtain from the manufacturers similar spectral curves for the camera including its optics. Thus, we must rely on the students' or our judgements concerning the intensity of the light or the intensity of the image on the video screen. These conclusions seem to be too subjective for serious measurement at this time.

An Excursion Into Chaos

We have attempted to use digital video to analyze the motion of a nonlinear, driven torsion pendulum. This pendulum, which was created for another project and which received an award from the American Association of Physics Teachers, can show a variety of motions and effects which are related to chaotic behavior. We had hoped that the digital video measurement tools, which we had created, would be a valuable way for students to visualize some of the details of the motion of this pendulum. For this type of motion the tool which we have most readily available is the "point and click" tool described above. At this time the use of that tool is much more tedious than using a more standard microcomputer based laboratory interface to collect data. The primary reason for this conclusion is the large quantity of data which one must collect. Using the point and click method is too time consuming for the students and causes frustration before they are able to obtain the necessary data. We believe that this situation will change when we build a little more flexibility as described above into our visual space-time diagrams. Thus, we will return to this problem in the near future.

High Speed Videography

While high-speed shutters up to 1/10,000th of a second are becoming common on consumer-level video cameras, high-speed frame rates are still financially far beyond the reach of most teaching laboratories. Thus, students who wish to record events which happened rapidly in comparison to 1/30th of a second must resort to either high-speed 16- millimeter film or high-speed strobe photography. Loren Winters, a high school teacher in North Carolina and a graduate of Kansas State University, has developed a method of using still camera strobe flash units with a video frame grabber. In his system a computer controls both the sequencing of several flash units and a video frame grabber. We are investigating whether a similar technique can be used with the video capture boards which have recently come on the market. At this time it is too early to draw a conclusion.

Conceptual Development in Modern Physics

With additional support of an NSF Graduate Fellowship we have augmented the development of materials in modern physics by investigating conceptual understanding of modern physics topics. The research has been divided into the study of two distinct groups: physics majors/faculty and non-physics majors. Several methods have been used with each group to enable them to demonstrate their knowledge and to improve our methods of investigating it. Through this research, we hope to improve teaching methods by uncovering differences among various types of students, conceptual difficulties which students have with modern physics, and students' misunderstandings of the concepts.

In the past year, we have run two major studies -- one for each group. The first study compared the knowledge of a cross section of physics majors, from freshmen through faculty members. The second study involved an in-depth study of the development of knowledge in non-physics majors who were taking a course in modern physics.

Since the beginnings of our studies, we have been interested in comparing the conceptual understanding of physics majors at different levels of education in modern physics. We spent the first two years testing and refining tools which probe students' conceptual knowledge. In September, 1993, we had developed a tool to measure this understanding. The tool, the Relationship Ranking Task, is based on the semantic network theory of memory which has been developed and tested by psychology and education researchers. This theory hypothesizes that the mind organizes ideas as a semantic network (or concept map) which consists of concepts linked together by their relationships. According to this theory, we can learn about a person's conceptual understanding by asking him/her to map concepts and the relationships joining them. To create the equivalent of a concept map the Relationship Ranking Task presents the subject with concepts in pairs and ask him/her questions about the relationship between the two concepts. During our study, we used a computer to present the physics majors with over 350 unique pairs of concepts from modern physics. For each pair of concepts, we asked the students to rank on a scale of 1 to 5 the relationship between the concepts on three different scales--the type of relationship, the strength of the relationship, and the importance of the relationship to physics.

During September and October, 1993, 49 physics majors (graduate and undergraduate) completed this task. Faculty members completed a similar exercise during the Spring, 1993. We then compared the responses of all the participants to the responses of all the other participants, including both students and professors, in order to discover the similarities and differences among physicists and students of physics at various educational levels.

The results were ordered based on education level, and graphs were drawn to show the similarity in response. A general trend was quite pronounced in the resulting data. The students from first year undergraduate through first year graduate students responded similarly. A somewhat high similarity also existed among third year graduate students through faculty members. These two groups differed markedly from each other as well. The second year graduate students seemed to form a transition between the two groups. From these results we may conclude that, as far as modern physics topics at Kansas State University goes, little conceptual change happens during undergraduate study or after the second year of graduate school. However a significant change occurs during early years in graduate school. Most graduate students at KSU were not undergraduate students here. Thus, the grouping of first year graduates with the undergraduates indicates that the result may be generalizable. Further research is planned to look further into what change happens during early graduate studies and why the change takes place when it does.

Parallel to the physics major studies, we have also been studying the modern physics conceptual development of some non-physics majors. This study revolved around the one-semester modern physics described above. To track the changing knowledge of the students during the semester, we used extensive concept mapping. Periodically throughout the course (often around exam time), we would require the students to draw a concept map of what they had learned so far. These maps were entered into a computer and compared in terms of the number of concepts present, the average number of relationships per concept, and the general complexity of the map. We have found that the maps are often as unique as the people who write them, which sometimes makes comparisons difficult. For this coming spring semester, we hope to investigate the effect of using a computer program to draw the maps, as opposed to hand-drawing. We have written a program for this purpose which will allow the students to rearrange and edit the concepts as they go. We also may incorporate some of the Relationship Ranking Task idea by having the students give the concepts and relationships a rank as they enter them.

QuickTime, Video for Windows, ActionMedia II, and All That

When we began this project the only relatively low-cost form of digital video which was available on a personal computer was Digital Video-Interactive which later had its name changed to ActionMedia II. At that time low-cost was a few thousand dollars. During the time that this project has been active, digital video has become available in several other formats for personal computers and the cost has dropped to a few hundred dollars. The most common formats are QuickTime, a product of Apple Computer, Video for Windows, a product of Microsoft, and Ultimedia, a product of IBM. Unlike ActionMedia II, each of these products can be played on computers which have no special video hardware. However, the quality of video for each of these is somewhat lower than ActionMedia II. (However, Ultimedia running under OS\2 2.1 is coming close to meeting the ActionMedia II standard.)

While the quality is not as good as ActionMedia II, the cost for video capture hardware and improved video playback is much lower. Thus, QuickTime and Video for Windows are becoming the digital video formats of choice for many physics teachers. Unfortunately, the nature of the hardware and software used for QuickTime and Video for Windows is somewhat different from ActionMedia II. Thus, more than a trivial amount of programming is needed to convert our materials into the more popular but lower quality video formats. On the other hand, the ideas of using video processing and image capture as we have described them are easily transportable. Thus, we will continue during the last few months of this project to investigate which items can be most easily converted to one of the popular formats.

Additional Support

Throughout the course of this project we have been fortunate to receive financial and in-kind support in addition to the National Science Foundation Grant. In the early days of the grant IBM provided computers and an ActionMedia Board Set. The graduate student who is investigating conceptual development in modern physics in support of our efforts has been funded by an NSF Graduate Fellowship. This past summer an undergraduate fellow received support from the Howard Hughes Medical Foundation. Finally, in addition to the matching funds which KSU committed as part of the original grant, it is now providing access at no cost to the project to high-end graphics workstations for some of our visualization work.

Appendices