FEd Fall 2001 Newsletter - The Evolution of Web-Based Activities in Physics at Illinois

Fall 2001



Previous Newsletters

Current Issue

Contact the Editor

The Evolution of Web-Based Activities in Physics at Illinois

Tim Stelzer and Gary Gladding

Five years ago, the Department of Physics at the University of Illinois undertook a complete revision of the introductory courses. The details of this revision have been reported in the paper "Parallel parking an aircraft carrier" in the FEd newsletter, Summer 1997. A key component of the course revisions was the implementation of web-based homework and an on-line grade book system. In this article, we will describe the evolution of this system to include delayed feedback homework, "Interactive Examples" with sophisticated help sequences, and preflights for "Just-In-Time Teaching."

The Department of Physics has been using computer-based homework for more than 25 years1. As part of our course revisions, we exported our Novanet-based exercises and grade book to the web. This move dramatically increased accessibility of our material for both students and faculty, and it has been very well received.

The original homework problems, which are still being used, typically consist of a physical situation about which the students answer several quantitative questions. A help button is also available which reveals hints to the problem? solution if requested. For example one problem reads: Two electrons and two protons located at the corners of a square as shown in the figure. Calculate the x and y components of the electric field at the center of the square, and the x and y components of the force on the left proton due to the other three charges.

The student can enter their answers in a textbox and hit the submit key. The computer immediately indicates which responses are correct and which are wrong. The student can then rework the incorrect responses and resubmit them, repeating the process as many times as necessary until all the answers are correct.

Our experience with this type of homework is similar to that chronicled at North Carolina State2. Students enjoy the flexibility of doing the homework from any web browser. They also appreciate the immediate feedback and typically continue with a problem until they have everything correct. Indeed, student scores on the web-based homework are very impressive. Often students come to office hours and say "I know the x-component of the electric field is supposed to be zero, but why?" This illustrates one of the strengths of immediate feedback. Students are provided with the opportunity to immediately identify their misunderstandings while the problem is fresh in their minds which greatly increases the incentive to confront their difficulties and clarify their understanding.

A disadvantage to immediate feedback is that it eliminates the incentive for students to check their own work. Students find it is much more efficient to attempt a problem and let the computer tell them if they're right. Indeed, a common mode of operation is for students to enter "0" for all of the questions and submit their answer. Since frequently several components of a vector are indeed zero, the computer will immediately identify these as correct, eliminating the need for students to think about these questions. They then focus their efforts on the "real" problems. To reinstate the incentive for students to check their work, we added one problem to each assignment that has delayed feedback. Students are permitted to resubmit and change their answer to this question as many times as desired up to the deadline. But, similar to hand-graded homework, they don't receive any feedback about the correctness of their answer until after the deadline for making changes has passed. These new delayed-feedback problems seem to have had the desired effect of forcing students to thoughtfully and carefully check their work, and we have observed a strong correlation between students' performance on these questions and their performance on exams.

Another common experience we noticed with students' interaction with web-based homework is that of students often asking for the help to be displayed before they even read the problem! As a result, the "helps" were being viewed as part of the problem statement. Indeed, we believe an important shortcoming of most computer homework is that the help and feedback mechanisms are typically a monologue, despite the educational research findings that a dialogue is more effective. To help address this shortcoming, we have developed web-based exercises called Interactive Examples (IEs) that are designed to actively engage the students in a Socratic dialogue and promote concept-based problem solving. In each of these IEs, students are asked a single, somewhat challenging quantitative question. If the student can successfully answer this question, credit is given for the exercise and some optional "follow-up" conceptual questions are asked to test this understanding. If the student cannot successfully answer the initial question, a help dialogue, which takes the form of a series of additional questions, some conceptual and some quantitative, guides the student to develop a problem-solving strategy to answer the initial quantitative question. Eventually, as the student makes use of the help given in the computer responses, he or she is able to arrive at a correct solution to the initial question. The amount of help needed varies with the student. The software is also designed to allow the student to interrupt the dialogue and answer the initial question at any time, thereby reducing the tediousness often associated with web-based homework questions. The student can also continue to ask for help and be led to deeper and deeper levels of interim questions. Once the student has successfully answered the initial question, a recap of the strategy is presented, and finally, optional follow-up questions are posed to allow the new knowledge to be tested and applied. We have developed over 50 IEs for use in our introductory courses, and you may view them at http://www.physics.uiuc.edu/tycho/index.html.

Students have received these new IEs enthusiastically. We have conducted informal interviews and anonymous surveys, and an independent team from our Office of Instructional Resources has conducted focus groups to assess student satisfaction. The reports are all consistent. Students find IEs intuitive and important to their understanding of physics. A typical comment from a student is: "it actually helps you when you're stuck to not only get the problem right, but helps out knowing how to do the same thing again and feeling confident in the physics behind the problem. This is a must. it would definitely help me do better in this class. it's like having a personal TA to assist you with every problem when you get stuck."

Another nice feature of IEs is that every submission a student makes while working through the problem is logged. We have used this information to study how students interact with the IEs, and our preliminary analysis is promising. Most students ask for some help initially, but are able to solve the problem without using all of the help available. As we continue to study these logs, we will continue to refine the IEs and assess their effectiveness.

All of our homework and grade book utilities are written as Perl scripts in a package we call Tycho. We continue to adapt these drivers based on our experiences as well as advancements in physics education research. Most recently, we modified Tycho to accommodate the introduction of preflights for Just-In-Time Teaching3 into our courses.

Preflights consist of multiple-choice and text box questions that students must answer prior to each lecture. These preflights encourage students to preview the material before lecture and also provide an opportunity for the lecturer to identify student difficulties with the material. In order to assist the lecturer in efficiently extracting information from the students' responses, we have incorporated a sophisticated preflight module into our grade book. In addition to providing statistics for each of the multiple-choice questions, the module also offers powerful filtering options to quickly identify common student difficulties. The faculty has been very pleased with this functionality, and preflights are now being implemented in several of our advanced, as well as introductory, courses.

Through our experiences of improving the introductory courses at Illinois, we have learned several important lessons. First, developing quality materials always requires a significant investment of both time and money. It is imperative that we combine our experiences and resources in this endeavor. Whenever possible, we have borrowed material directly from, or based our work on ideas from, the physics education community. In a similar spirit, we encourage others to take advantage of our experiences and materials and assimilate them into their courses. You may view all of our work at www.physics.uiuc.edu/tycho/index.cfm, or contact us at tycho@uiuc.edu.


  1. D.J. Kane and B. Sherwood, "A Computer Based Course in Classical Mechanics" Comp. and Educ. 4, 15-36 (1980); L.M. Jones, D.J. Kane, B.A. Sherwood and R.A. Avner, "A final exam comparison using computer based instruction," Am. J. Phys. 51, 533-538 (1983); L.M. Jones and D.J. Kane, "Network use in central management of large university physics courses," J. Comput.-Based Instr. 19, 77-81 (1992); L.M. Jones and D.J. Kane, "Student evaluation of computer-based instruction in a large university mechanics course," Am. J. Phys. 62, 832-836 (1994).
  2. Scott Bonham, Robert Beichner and Dueane Deardorf "Online Homework: Does It Make A Difference?" Physics Teacher 39, 293-296 (2001).
  3. Gregor M. Novak, Evelyn T. Patterson, Andrew D. Gavrin, and Wolfgang Christian, Just-in-Time Teaching: Blending Active Learning with Web Technology (Prentice Hall, Upper Saddle River, NJ, 1999).

Tim Stelzer and Gary Gladding are in the Department of Physics at the University of Illinois (Urbana-Champaign). Tim is a researcher in the physics education group, and is currently working on the development and evaluation of interactive examples. Gary is an experimental high-energy physicist who has devoted a significant fraction of his time to physics education in the last five years, and led the introductory course revision effort at Illinois.