Skip Office Navigation
OPE: Office of Postsecondary Education
Current Section

Archived Information

Lessons Learned from FIPSE Projects II - September 1993

University of Oregon

Biology Laboratory Construction Kit With Intelligent Tutor


Previous computer simulations used for instruction in introductory biology have employed models of a single fixed biologic system, for example, simulations in physiology for medical students. This focus on particular systems did not permit introductory-level study of cardio-respiratory systems across species, phyla or kingdoms.

The purpose of this project was to overcome these limitations by developing a computer toolkit and intelligent tutor that would place students in a flexible computer environment for learning basic concepts of a variety of circulatory systems. Students generate hypotheses, design experiments, test realistic simulations and learn how their constructions meet or fail to meet design objectives. The program invites investigative thinking, and particularly develops the ability to draw conclusions and unravel causality.

Innovative Features

The project attempted to connect the best of the constructivist approach to the use of computers--the cardiovascular toolkit--with the best of the tutorial--the intelligent tutor. The toolkit (programmed in Allegro Common Lisp on MacIntosh computers) allows students to construct a cardiovascular system graphically by hooking together appropriate parts. The tutor was designed to guide and critique student work. Students start by selecting cardiovascular components (hearts, vessels, capillaries) as graphic objects from a menu of icons. Gauges capable of measuring blood pressure, flow, volume and oxygen can be attached to the components to visualize the system's status. Components and gauges can be moved around and hooked together to build new structures. Parameters for each component, such as resistance to flow or pumping rate, can be easily modified to create changes in cardiovascular behavior.

Thus, by selecting and placing graphic objects on the computer screen, the components can be connected in a number of topological ways. The user can then run the system by clicking on a control panel, and can, if he or she chooses, accelerate or animate heart beats, blood flow and one-way value operation. Graphing windows associated with gauges immediately display factors such as pressure, flow, and oxygen content.

The curriculum built around the tutor is designed for two-hour lab periods, supported by lecture and pre- and post-lab discussion sessions. It first acquaints students with model behavior and instructs them on how to pose problems and design experiments. In later laboratories, students devise their own hypotheses, design their own experiments, and practice drawing conclusions based on results. The software tutor helps instructors coach, explain or critique students as they are making constructions.


Student and instructor curricular materials were tested in small groups, using four experiments on the cardiovascular system. A second lab evaluation tested students' ability to generate hypotheses and draw logical conclusions. A third tested students' skill at maximizing the performance of a construction for producing, transporting, and consuming oxygen.

Project staff used videotapes of pairs of students using the lab instructions and simulations to assess the adequacy of the software interface, the instructional design, and prior student knowledge, abilities and misconceptions. Extensive testing during lab experiments identified key problems in both the students' biological understandings and in the human-machine interface. The students' misconceptions turned out to be deeply held and related to basic principles of physics. Fortunately, the videotapes showed that the software helped eliminate the misconceptions, as students discovered for themselves the flaws in their logic and in their assumptions. Both the videotapes and classroom testing showed that students and faculty using the software found it instructive and easy to use. In fact, most students learned to use the program without the manual.

The Construction Kit and curriculum were distributed to 18 additional sites for field testing (universities, colleges, community colleges and schools) as part of the testing of BioQuest software (Quality Undergraduate Educational Simulations and Tools in Biology, Beloit College). Although BioQuest requested evaluations of the software from all of its test sites, especially on student use and learning, only a few sites responded. Most comments from field sites were related to problems in the software (especially regarding compatibility problems with various versions of the MacIntosh system).

Project Impact

The Construction Kit and curriculum are being used by over 500 students each year in at least four different biology courses at the University of Oregon. The software has become part of a larger curricular project (funded by FIPSE and the National Science Foundation) designed to encourage investigative learning in biology courses for non-majors.

Unanticipated Problems

Developing causal explanations for the behavior of arbitrary constructions proved to be more complex for the faculty than anticipated. Also, students' false expectations confounded interpretation of test data. Definitional problems arose out of the testing of "normal" versus "abnormal" physical properties and of absolute versus relative measurement of these properties.

Although significant progress was made in developing the intelligent tutor, it proved unrealistic to incorporate the tutoring software into the final software product within the time-frame of the project.

Major Insights And Lessons Learned

Faculty often had difficulty pinpointing the nature of student problems simply from observing them using the software. When a student felt the simulation had behaved unexpectedly, for example, there were at least four explanations: 1) the student did not understand how the simulated system represented the real systems being simulated (a problem with the interface); 2) the simulated system did not accurately represent the real system (a problem with the simulation model itself); 3) the student had misconceptions about the real system, and/or 4) the student was misinterpreting the results.

The first two explanations complicate the use of simulations in education and underscore the need for thorough testing of both the user interface and the simulation model. If a simulation is to erase student misconceptions about real biological systems, then both students and instructors must have confidence that what they are seeing in the simulation accurately reflects what happens in real life. Otherwise, it is not possible to know whether to focus on the student or on the software when something unexpected happens. These problems were especially acute for this software because of its ability to design systems of arbitrary complexity, challenging even expert faculty to know what to expect.

Generally, designing and building effective educational software is very difficult, time-consuming and expensive. Testing the pedagogical goals of the software while it is being developed and while it is in use is essential, but very little software is subjected to this kind of review.

Project Continuation

The software is used in several University of Oregon biology courses. Testing at the 18 sites continues; therefore, neither the software nor the curricular materials are quite ready for final publication. When FIPSE support ended, the University, Apple Computer, and BioQUEST funded the last stages of the project. The software will be published on CD-ROM with other BioQUEST materials in 1993 by the University of Maryland Press.

One of the project directors, the Chair of the Biology Department, was awarded a second FIPSE grant of $257,000 to support workshop biology for non-majors. This project, which promotes scientific literacy through investigative labs and issue-oriented activities, was influenced by the student-directed and open-ended investigations of the Construction Kit. Recently, the toolkit has been modified to be more useful in a workshop format.

Available Information

Beta versions of the software (the Cardiovascular Construction Kit) and the lab and users' manuals are available from the project directors:

Daniel Udovic
Department of Biology
University of Oregon-POB 3158
Eugene, OR 97403

Nils Peterson
From the Heart Software
Eugene, OR 97403

Sarah A. Douglas
Department of Computer and Information Science
University of Oregon
Eugene, OR 97403

The software runs on MacIntosh computers. It requires a hard disk and at least four megabytes of memory. Please send two empty 800-kilobyte or high-density diskettes. In return, the project directors ask for help in evaluating the software and curricular materials at additional campus sites.

The Intelligent Tutor is also available, as is a 17-minute videotape of the toolkit in use.

The software and manuals are also being distributed as part of the BioQUEST project. BioQuest simulations are published on CD-ROM by the University of Maryland Press. For information about BioQUEST, write to:

John Jungck
Project BioQUEST
Beloit College
Beloit, WI

Several articles based on the project include:

Douglas, S., Peterson, N.S., and Udovic, D. "The Cardiovascular Construction Kit." In BioQUEST: Quality Undergraduate Educational Simulations and Tools, University of Maryland Press. 1993.

Douglas, S.A. and Zheng-Yang Liu. "Generating Causal Explanation from a Cardiovascular Simulation" in Proceedings of the International Joint Conference on Artificial Intelligence, Detroit, MI, August 1989.

Downing, Keith Linn. 1990. The Use of Teleology in the Qualitative Evaluation and Explanation of Circulatory Systems. Ph.D. Dissertation. University of Oregon. (Available through University Microfilms.)

Novak, David. "A Laboratory Curriculum for the Cardiovascular Toolkit". Thesis submitted in partial completion of the Master's Degree, Summer 1989.

[Northwestern University] [Table of Contents] [University of Rhode Island]



Last Modified: 02/22/2006