ORMS Today
August 1999

Teaching With the Lights Out

By Kenneth N. McKay


In 1991 Stephen Graves (MIT Sloan School), David Kletter (then a student at MIT) and I decided to create a rapid software development system for the types of models found in operations research and operations management. The system was designed to allow us to develop initial prototype educational modules in days, not weeks or months. By the fall of 1993, the first version of the system was ready for testing, and we built seven modules. In February 1994, the lights in my class were turned out, and the students were not the only ones to wonder what would happen next.

Consider what it means to teach with the lights out. The room is dim and everything is projected from the beginning of the lecture to the end. The lecture notes are on the computer and are shown with the necessary text and mathematics. The computer system has to facilitate the primary presentation or exploration of the material and has to support the secondary experiences. For example, in a lecture on safety-stock inventory, the basic concept of inventory reduction and replenishment must be understood, the mathematics presented, and possibly a simulation shown to illustrate the phenomena. Many instructors use software such as MS PowerPoint or Excel for this "primary delivery."

However, learning and interaction are enhanced by running "what-ifs," answering students' queries, and getting the students involved with the models. To achieve this, the instructor may need complete control over the user interaction and model execution. Some of the control and interaction can be achieved using spreadsheets, but our software overcomes some limitations of spreadsheets. For example, in guided instruction you may need to make various aspects protected (e.g., all but one coefficient in an equation), provide feedback when a certain data value is clicked on a chart, or animate part of a chart or model. Whereas spreadsheets are designed for "getting the job done," we designed our software for teaching and attempted to address the types of interaction and control needed in the classroom or in self-guided modules.

Imagine what you could do with a remote mouse control, a laser pointer and software designed to explore the models and support what-if questions from the students. Imagine random problem generators and computer quizzes, or software that animates algorithms and steps through them one iteration at a time. You are not restricted to staying at the front of the room; blackboards are not needed. You can effectively and efficiently show the differences between two options. You can ask students what they think will happen next before you press the proceed button. The students can do the steps and then see the answers presented on the screen. The students can see interactive sensitivity analysis and discuss how a specific coefficient contributes to the behavior of the model.

Imagine the changes to your class. Students learning in the classroom — quite possible. The same lecture year-in year-out — extinct. The comfortable feeling of knowing your one example inside out — gone.

In the computer-facilitated classroom with exploratory software, the students drive the class and discussion. The what-ifs are different from class to class. What you are asked to illustrate and explain is different. Imagine all your quantitatively oriented lectures like this, not just one. It brings the models alive and makes the classroom a dynamic place. Students who want to memorize the answer for the final will not enjoy the experience. Students who want to learn and understand will.

Needless to say, the pedagogical style (and the software) has evolved since the first lectures in 1994. Here are some of the lessons learned:
  • Different topics require different tools, but there are common concepts of calculators, explorers, algorithm steppers, random problem generators and quizzes.

  • Software used in the classroom must be fully open and unrestricted to allow the instructor to quickly craft examples and access any aspect of the model whereas the student or lab version may be guided and restrictive for computer-aided-learning.

  • Encouraging students to help each other during the class participation periods is an excellent technique for learning and getting the students involved.

  • It is important to show "wrong" answers to illustrate what errors look like.

  • The ability to hide parts of any graph or table and then slowly uncover them is important for engaging students in dialog.

  • Anything with an algorithm must be capable of being "stepped" forward to show how an algorithm progresses and then backward to show again and again a key point.

  • The use of color should be minimized and if used, additional visual cues are needed for the color-blind. We learned this the hard way.

  • It should always be obvious where you are within the software, what can be done, what options are active, and what is expected next.
It has now been eight years since we started this project. The students love the visual what-if capability for subjects like forecasting, inventory and statistical process control. After teaching this way, it is impossible to go back to overheads, foils and chalk.

In 1998 we re-created the development tool in the Java programming language and then used it to create an initial suite of teaching modules for OR/MS. These second-generation modules are different from the first in many ways, but the intent is the same: exploring the behavior and relationships in the models. The first versions of the applets can be found at the "Interactive Learning Center" which is part of the MIT Operations Research Center's Web site: http://web.mit.edu/org.



Kenneth N. McKay is an associate professor in the Faculty of Business Administration at the Memorial University of Newfoundland. He can be reached by E-mail at kenmckay@morgan.ucs.mun.ca.





  • Table of Contents

  • OR/MS Today Home Page


    OR/MS Today copyright © 1999 by the Institute for Operations Research and the Management Sciences. All rights reserved.


    Lionheart Publishing, Inc.
    506 Roswell Street, Suite 220, Marietta, GA 30060, USA
    Phone: 770-431-0867 | Fax: 770-432-6969
    E-mail: lpi@lionhrtpub.com
    URL: http://www.lionhrtpub.com


    Web Site © Copyright 1999 by Lionheart Publishing, Inc. All rights reserved.