DESIGNING COMPUTER SIMULATED EXPERIMENTS TO OVERCOME STUDENTS' MISCONCEPTIONS ABOUT DISPLACEMENT OF LIQUIDS

by Toh Seong Chong Centre for Educational Technology & Media Universiti Sains Malaysia 11800 Penang, Malysia e-mail: tohsc@usm.my

Abstract
This study compared the effectiveness of computer simulated experiments with that of parallel instruction involving hands-on laboratory experiments for teaching volume displacement concepts. The purpose of the simulation was to have students test their misconceptions rather than simply being told about erroneous misconceptions. The theoretical underpinnings and instructional design model used in the design of the simulation courseware were discussed. This study also assessed the differential effect of students' understanding of the volume displacement concepts in the cognitive category of knowledge and application. In addition, it compared the degree of retention, after 30 days, of both treatment groups. 389 students from 6 Malaysian schools participated in the study. The results indicated that there were significant differences between the two groups in terms of learning gains in the cognitive category of knowledge as well as in the cognitive category of application. The implications of these results for the design of computer-based instruction for the teaching of science are discussed.
*Paper presented at EDUCOMP 96:Tomorrows Education Today sponsored by the Malaysian Council for Computers-in-Education held at Putra World Trade Centre, Kuala Lumpur, 28-29 May 1996. Mr. Toh Seong Chong is a lecturer in Instructional Design and Multimedia at the Centre for Educational Technology and Media, Universiti Sains Malaysia.

Toh Seong Chong

EDUCOMP96

Introduction
The advent of microcomputers has resulted in an increased interest in the design and use of instructional simulations. It has made possible a new and exciting form of learning environment (Reigeluth & Schwartz, 1989). This type of simulation has the potential to enhance the transfer of learning by teaching complex mental and procedural tasks in an environment that approximates a real-world setting. It provides the student with an environment to build a useful mental model of part of the world and gives him an opportunity to test it safely and efficiently (Alessi & Trollip, 1991). However, the microcomputer-based simulations being produced today are often poorly designed from a psychological standpoint, and the learning environment they provide is ineffective at best. Information concerning how to specify and implement appropriate instructional designs for this type of learning medium is sparse (Alesei, 1988). The purpose of this research is both developmental and experimental in nature. Specifically, the purpose is to describe, construct, and test a series of learner-based computer simulations based on instructional design principles to teach students concepts pertaining to displacement of liquids.

Theoretical Framework
The underlying theory base for this prescriptive model for the design of computer simulation comes from "the Model of Information Processing Theory of Learning and Memory" by Gagn & Briggs (1979). The Model of Information Processing Theory of Learning and Memory originated from the work of Gagn (1977) and Bransford (1979). This model sees the mind as an information processor; doing the things that a computer does. The mind accepts and encodes information, manipulates and transform, stores, and retrieves it. Memory in this model is an active mental process by which information is coded, stored, retrieved and integrated with previous information. Memory is divided into three systems namely the sensory memory, short-term memory and long-term memory. The sensory memory has two important characteristics: it can store information for only a brief time, about 0.075 of a second, and it can store an enormous amount of information. It is then dropped or transferred into the short-term memory. The control process that governs what information will go from sensory to short-term memory is attention. The short-term memory is the "working memory". It is where information is consciously thought about and manipulated. Information remains in the STM for less than 20 seconds. Miller (1956) showed that the STM can hold only seven bits of information, plus or minus two. At the end of that time, information is forgotten or transferred to the long-term memory. Reorganising discrete items into classes, categories or chunks can enlarge the capacity of the STM. This process is called chunking. The duration of the information in the short-term memory can be sustained for a longer period of time by giving it continued conscious attention. This process is called rehearsal. Rehearsal is also the mechanism by which information is transferred to the long-term memory. The long-term memory is quite different from the other two parts of our memory system. Its capacity is almost unlimited, and once embedded in it, the information stays forever. The process of long-term memory storage is called semantic encoding because what is stored is not

Toh Seong Chong

EDUCOMP96

the information itself but some more efficient verbal representation of it. The information is transformed into meaningfully and purposefully connected verbal chunks that have been referred to as semantic networks (E. D. Gagn, 1985).

Design of the Computer Simulations

Based on the Model of Information Processing Theory of Learning and Memory, a series of simulations were designed. The actual instructional design process was based on the Events of Instruction by Gagn and Briggs (1979). The nine sets of instructional external events were arranged so as to ensure that the desired learning process took place internally. Detailed design guidelines of the simulation were based on suggestions from an instructional design template suggested by Reigeluth and Schwartz (1989) which is sketched in Figure 1.

Instructional layer Scenario Model

Fig. 1 General simulation program structure A simulation is seen as a series of layers starting with a core which is the model. Around this core is another layer called the scenario, and finally we have the instructional layer. The model reflects the relationships which govern the situation being modelled. The scenario describes that situation (what happens and how, who is involved, etc.). The instructional layer tries to provide as much as possible an individualized and effective learning experience to each learner. The first version of the simulation was written by the author using using Turbo C, (Version 2) as the programming language. The graphic and animation screens were designed using IBM P.C Storyboard Plus and later linked to the Turbo C compiler. To conserve diskette space taken by the graphic images, a routine was developed to compress the graphic images from 16 Kilobytes to 3 kilobytes and stored; then subsequently decompressed and read when required.The second version of the simulation was developed by the author using object-oriented programming (OOP) as found in the Asymetrix Toolbook (Ver 3.0) Multimedia Authoring Tool. The design of the simulation template consists of three components namely the Model Editor, the Educational Scenario Editor and the Educational Instruction Editor.

Toh Seong Chong

EDUCOMP96

The Model Editor allows specification and calculation of simulation models. The Educational Scenario Editor allows the development of a scenario for a simulation model including use of pictures, different forms of animation (movement, change in colour, change in number, flash effects, etc.), other forms of representation (numeric, graphic, tabular, iconic, etc.) and change of parameters. The Educational Instruction Editor allows the construction of an instructional layer to encapsulate the whole. The instructional layer involves the questions posed, the analysis of the answers given, feedback and help, the role of the learner and the strategy used. It is intended to allow different learning styles (exploratory, guided discovery, etc.). The nine events of instructions as suggested by Gagn and Briggs (1979 were incorporated into the three components of the simulation. A typical simulation was presented in the following manner: Events of Instruction 1. Gaining attention Procedure 1. Present a graphic screen with animation and sound to attract the students' attention. See Fig. 2

Fig 2. Screen to attract attention 2. Informing learnerof lesson objectives 2. Display of menu screen that showed summary of concepts. See Figure 3.
MAIN MENU DISPLACEMENT OF LIQUIDS 1. Volume of Object 2. Size of Object 3. Floating objects 4. Objects broken down 5. Shapes of Object 6. Types of Liquids Use direction keys to choose, F3 for help, F4 to exit Press <RETURN> when done

3. Stimulating recall

Fig. 3. Screen informaing user lesson objectives 3. Display bullet text that is accompanied with sound to assist student recall, i.e. explain purpose of simulation. See Fig 4.

Toh Seong Chong

EDUCOMP96

SIMULATION 1.

VOLUME OF OBJECTS

In this simulation, you will be given two objects that have the same volume, but different mass.When each object is immersed in the eureka can, water is displaced. We will see if the volume of water that displaced depends on the mass or the volume of the object. Press < ENTER > TO CONTINUE

4. Presenting scenario

4. Simulating two objects A and B that have the same volume but different mass being dropped into the eureka can separately.See Fig 5.

Fig. 5. Presenting scrnario

5. Guiding learning through 5. Simulating water flowing out of the eureka can and being simulation model measured using a measuring cylinder. The volume of water displaced depends on the volume object, not its mass. See Fig.

Toh Seong Chong

EDUCOMP96

6. Eliciting performance

6. After user has chosen the object he wants, he has to press the <ENTER> key. If the wrong key is pressed a "pop-up" message is displayed accompanied by a warning sound. See Fig. 7

Fig. 7. Eliciting performance

7.

Providing

informative 7. Displaying a graphic screen line by line followed by sound to

Toh Seong Chong

EDUCOMP96

feedback

give the user a summary of the concept that he has just learnt. See Fig. 8

8. Assessing performance

8. Giving a quiz to assess user's understanding of the concept. See Fig 9.

Fig. 9. Assesing performance

9. Enhancing retention

9. Using the branching technique, if the learner could not give

Toh Seong Chong

EDUCOMP96

the correct response after two trials, he will be guided to the simulation again. See Fig. 10

Toh Seong Chong

EDUCOMP96

Procedure
An experimental design using a control group known as "The Pretest Posttest Control Group Design 2 " (Campbell and Stanley, 1963) was used in this study. The dependent variables were gain in scores between the pretest and posttest and the retention test scores. The independent variables were treatment, cognitive category in the test and gender of students. Students in each class were randomly assigned to one of two treatment groups, the microcomputer-simulated experiment group (the experimental group) and the hands-on laboratory experiment group (the control group). The experimental group consisted of 97 male and 91 female students. The control group consisted of 94 male and 107 female students. The experimental and control groups of students were both taught six important volume displacement concepts. The Experimental Group The experimental group was taught these six concepts using a series of six simulated experiments on the IBM PC microcomputer. The simulation courseware consisted of six simulated eperiments. Six simulated experiments were included in the ourseware. The peer tutoring and learning approach was used where three students shared a computer. Each of the students took turns to become the group leader in each simulation. The Control Group The control group was taught these concepts using six parallel hands-on laboratory experiments by several science teachers. The experiments were parallel to those used in the microcomputer simulated experiments. Instruments There were four instruments in this study. They included the Cattell Culture Fair Intelligence test and the Physics Achievement test which was subdivided into the pretest, posttest and retention test. The Physics achievement test consisted of a twenty four multiple choice questions, half of which were from the cognitive category of knowledge while the other half were on the cognitive category of application. These questions were developed to determine the students' understanding of important concepts related to volume displacement. An item analysis was carried out on the results of the testing and employed three types of information - item difficulties, discrimination indices and the pattern of responses to the various distracters - to improve the test. Content validity of the test was established by five science specialists who validated the instrument. Reliability of the test was estimated using the Cronbach Alfa procedure. The Cronbach Alfa coefficient was 0.89 showing that the test instrument was satisfactorily reliable. The Cattell Culture Fair Intelligence test was administered to all the students in this study to see whether randomness was achieved in the assignment of the subjects to the two groups in terms of intelligence. A pretest was administered before the treatment to both groups. Immediately after the treatment, a posttest was administered to the students. A retention test was administered to the students 30 days later.

Results
9

Toh Seong Chong

EDUCOMP96

The results were reported in Toh, Seong Chong (1993). A summary of the findings are provided here for discussion. 1. There is significant difference in the cognitive category of knowledge between the performance of students in the computer simulated experiment (CSE) and that of students in the hands-on laboratory groups (HOL) as measured by the gain scores between the posttest and the pretest. There is also significant difference between the performance in the cognitive category of application between the CSE group and the HOL group as measured by the gain score between the pretest and the posttest. There is significant difference in learning gains in the cognitive category of knowledge between the performance of males and that of females in each treatment group. Females showed higher overall gain score compared to males. The gain scores for the females were higher in both the CSE group as well as in the HOL group. There is no significant difference in learning gains in the cognitive category of application between the performance of males and that of females in each treatment group. There is significant difference in retention in the cognitive category of knowledge between the performance of students in the CSE group and that of students in the HOL groups as measured by a retention test. There is also significant difference in retention in the cognitive category of application between the performance of students in the CSE group and that of students in the HOL groups as measured by a retention test.

2.

3.

Discussion
The performance of the students in the computer simulated group had higher achievement compared to the hands-on laboratory group in the learning of volume displacement concepts supported studies done by other researchers such as Choi, B. S. & Gennaro, E. (1987) and Cavin & Lugowski (1978). Hence the teaching of concepts in science using computer simulations could be a viable alternative in lieu of hands-on laboratory experiments. This study also showed that there were differences in achievement between males and females; i.e. females seemed to score higher than males in the cognitive category of knowledge. This finding could be spurious and inconclusive. Up till now there is no concrete findings that suggested that there are gender differences in cognitive achievement. Maccoby & Jacklin (1974), concluded that gender differences existed for (a) verbal ability, (b) quantitative ability, and (c) spatial ability. But a meta-analysis conducted by Linn & Hyde (1989) on gender differences in cognitive achievement revealed that gender differences in these areas have declined and not uniform within these categories. Ruel (1990) suggested that gender differences in achievement could be due to maturation, which favoured girls over boys, cultural and sex-role expectations that precipitated differing student performances.

Conclusion

10

Toh Seong Chong

EDUCOMP96

This study is an attempt to formulate a prescriptive model for the design of computer simulations using the Model of Information Processing Theory of Learning and Memory. These prescribed strategies when tested showed that computer simulations can be used in place of a hands-on laboratory experience. Achievement of students using this mode of learning is comparable to that of hands-on laboratory experience. Hence the use of computer simulations to teach concepts in science seem to be a good alternative. Computer-simulations could also be a good substitute for expensive and dangerous experiments. Yet another implication of this study is that concept learning through computer-simulations speeds up the learning time. Students learning through the CSE achieve an equal performance level in approximately one fourth the time required for the hands-on laboratory experiments. Further research, based on good instructional design and better software tools should show even more significantly the efficacy of computer simulations as a medium of instruction.

References
Alessi, S. M. (1988). Fidelity in the design of instructional simulations. Journal of ComputerBased Instructions 15(2), 40 - 47. Alessi, S.M. & Trollip, S. (1991). Computer based instruction: Methods and development. (2nd ed.). New York: Prentice Hall. Bransford, J. D. (1979). Human cognition, Belmont, CA: Wodsworth. Campbell, D. T. & Stanley, J. C. (1963). Experimental and Quasi-Experimental Designs for Research. Chicago: Rand McNally College Publishing Co. Cavin, C. S., & Logowski, J.J. (1978). Laboratory experiments and student aptitude on achievement and time in a college general chemistry laboratory course. Journal of Research in Science Teaching, 15, 455-463. Choi, B. S. & Gennaro, E. (1987). The Effectiveness of Using Computer Simulated Experiments in Junior High Students' Understanding of Volume Displacement Concept. Journal of Research in Science Teaching, 24(6) 539 - 552. Gagn, E. D. (1985) The Cognitive Psychology of School Learning. Boston: Little Brown. Gagn, R. M. (1977). The Conditions of Learning, (3rd ed.). New York: Holt Rinehart and Winston, Inc. Gagn, R. M., Wager, W. & Rojas, A. (1981). Planning and Authoring Computer-Assisted Instructions Lessons, Educational Technology, 21(9), 17 - 26. Gagn, R. M., & Briggs, L. J. (1979). Principles of instructional design. (2nd ed.). New York: Holt, Rinehart & Winston, Inc.

11

Toh Seong Chong

EDUCOMP96

Linn, M. C. & Hyde, J. S. (1989). Gender, Mathematics and Science. Educational Researcher, 18(8), 17 - 27. Maccoby, E. & Jacklin, C. (1974). The Psychology of Sex Differences. California: Stanford University. Miller, G. A. (1956). The magic number system, plus or minus two: Some limits on our capability for processing information. Psychological Review, 63, 81 - 97. Reigeluth, C. M. & Schwartz, E. (1989). An Instructional Theory for The Design of Computer-based Simulations. Journal of Computer-Based Instruction, 16(1), 1 - 10. Ruel, A. A. (1990). Gender differences in spelling achievement in grades 1 through 6. Journal of Educational Research, 84(4). Toh, Seong Chong (1993). Use of Microcomputer Simulations to Overcome Students misconceptions about displacement of liquids. Paper presented at the Australian Association for Research in Education Annual Conference, Fremantle, W.A. 22 - 25 November, 1993

12