191

Research In Science Education, 1986, 16, 191-198.

USING COMPUTER SIMULATIONS TO PROMOTE

ACHIEVEMENT AND TRANSFER

Ronald Oliver and James Okey

Computer sire ulations have played an increasing role in computer-assisted instruction

(CAI) in recent years. Early CAI was devoted primarily to drill and practice but the

increased capability of school computers has been accompanied by more attention to

higher level objectives toward which most simulations are directed.

The role that simulations of all kinds have in instruction is to present some simplified

model of reality that can be observed and altered. The simulation provides a means of

experimenting with the world. Computer simulations are especially important because

they allow interaction with difficult, dangerous, expensive, or time-consuming events. A

student with a computer simulation can, for example, try wide ranging temperatures to

determine their effect on reactions or determine successive generations of fruit flies in a

matter of minutes.

In addition to the pedagogic value of computer simulations (i.e., compressing time,

reducing dangers, etc.), they promise some significant advantages in learning as well.

Because computer simulations allow students to make choices and observe and act on

consequences, they provide a way to study cause and effect relationships, to make

predictions, to test hypotheses, to gather data, and to draw conclusions. It is these

outcomes that constitute higher level thinking and problem solving that are frequently

mentioned among the important objectives of schools but not so frequently found in the

daily activities of classrooms.

Publishers, through the materials that accompany computer simulations, are not

timid in setting forth the high level objectives to which their simulations apply. For

example, one publisher of a simulation aimed at upper primary level students claims that

use of the simulation will increase reading comprehension, enhance communication,

promot e co-operative behaviour, and provide problem solving experience. A problem with

computer simulations such as this is not that the objectives are outlandish or that claims

are distorted or excessive. Rather the problem is one of face validity - what students do

while using the simulation doesn't involve obvious school tasks and thtts the link between

192

using the simulation and achieving the high level objectives is not clear. In this particular

simulation, students attempt to assemble a small raft on the screen so they can escape

drowning in an impending flood.

Because simulation tasks such as building rafts or managing buffalo herds are not

obvious school tasks, the value of the simulation must be sought in what it helps students

learn beyond the simulation itself. Thus the simulation skills themselves are not as

important as showing that they can be applied or transferred to other tasks of more

obvious value. The computer simulation becomes important not for what students do while

using it but rather what other things they can now do because of some generalizable skills

they have acquired.

Two major questions therefore arise when evaluating computer simulations that

provide somewhat fanciful tasks for students.

i. Can students successfully engage in the simulation tasks?

2. Can the simulation skills be transferred or applied in other situations?

PU R PO SE OF THE STUDY

The purpose of this study was twofold. First, means of acquiring skills using a

computer simultion were studied. The study alternatives were each examples of ways that

schools presently organize computer study. Secondly, the purpose was to see if the skills

acquired when using the computer simulation could be applied in other settings. In other

words, the second purpose looked at the transferability of the skills to new, non-computer

problems.

PROCEDURES

Sixth grade students (n=50) in three different schools used computer simulation in

three different ways. In the first school the students worked in pairs each sitting in front

of the terminal and alternatively taking turns in responding to the simultion tasks. At the

second school students worked in groups of four. Although a bit more crowded, they too

all sat in front of the terminal and entered responses to the simulation events. Students in

the third school were directed as pairs to work at the computer terminal as space and time

premitted. In this instance only a single computer was used over a number of school days

so that all students had an opportunity to interact with the computer simulation. In the

first two schools, the students were taken to a computing laboratory on a nearby campus

where all students in the class simultaneously worked on the simulation either in pairs or

groups of four.

The simulation used, called Raft Away River (Jacaranda Wiley, 1984), places

students in a river valley following an accident on a rafting trip. The accident leaves

them stranded with no way to safety unless they can build another raft from available

193

trees. During the process of building the raft they must feed themselves by catching and

cooking fish or eating berries. Periodic rains and bad food occasionally cause illness and

suspend work on the raft. Because each participant has specific tools (an axe, rope,

fishing line, or matches) they must cooperate in the cooking, eating, tree cutting, and raft

building. According to the literature provided with the computer program, it provides

students "an opportunity to develop skills in reading comprehension, communication,

cooperative behaviour, and problem solving."

Students in the three groups worked with the computer simulation over a period of

six weeks. The two groups that used the computers in the laboratory setting were each

scheduled for five sessions lasting 45 minutes to an hour. Approximately 20 minutes was

required at the first session to explain procedures and orient them to the computers but

after this all time was devoted to interactions with the simulation. Computer use by the

third group of students was considerably less systematic. A single computer was available

in the classroom for one half day each week. During this half day the teacher directed

pairs of students to use the computer. The sessions for these students were shorter

(usually about 15 minutes) and were less frequent (2 to 4 sessions for each student

compared to 5 for the other groups). In terms of total time for computing, students in this

latter group had approximately 25 to 30 percent as. much time working with the simulation

as students in the other two groups.

The total time for using the simulation was the same for each of the two groups

using the computer laboratory (five sessions of 45 to 60 minutes). However, there were

differences in use because of participating as a member of a pair or a foursome. Each

group would make essentially the same number of responses but the individuals in a pair

would have twice as many individual choices to make. Students in pairs and foursomes

also take on somewhat different roles because tasks can be more differentiated within a

larger group. For example, in a foursome one person can cut wood, another can carry it to

the raft building site or fire, another can be catching fish, and still another can be building

the raft. In a group of two each individual must take responsibility for more than one of

these chores.

Four measures were completed by each student - one prior to the computer work and

three afterward. The prior measure was a test to determine the logical thinking ability of

students. The three measures following use of the simulation measured attitudes toward

computing, knowledge of the procedures and skills involved in using the simulation, and

ability to transfer simulation skills to new topics. Each of these measures is described

more completely below and in Table I.

Group Assessment of Logical Thinking (GALT).

This 10-item multiple choice test requires double responses to each item - one

requesting an answer and the second a justification for it. The ten items are used to

assess five kinds of reasoning ability - correlational, probabalistic, conservation,

194

combinatoria, and proportional. This test has been shown to be both a valid and reliable

measure of logical thinking ability (Radrangka, Yeany & Padilla, 1983). Logical thinking

was thought to be a prerequisite for successfully completing the computer simulation in

this study.

TABLE 1

INSTRUMENT DATA

Possible Test Number Score Mean Standard Relia-

Measure Type of Items R a n g e S c o r e Errors bility

Group assess- Multiple ment of Choice Logical and Thinking Corn pletion

Attitudes Likert Toward Scale Computing

Raft Away Multiple River Choice,

Co m pletion, and Problem Solving

10 0-10 3.2 .24 .62

24 24-96 80.0 I.I .78

17 0-20 12.9 .49 .71

Transfer of Problem Learning Solving l0 0-18 8.0 .53 .73

Attitudes Towards Computing (ATC)

This is a 24-item Likert scale in which students are given statements (e.g. I find

computers difficult to use or Computers help you to learn) and asked to rate them as

Strongly Agree, Agree, Disagree, or Strongly Disagree. Items on the scale are designed to

asseas three dimensions of attitude in using computers to learn, and the value of

computers as aids to learning.

Raft Away River (RAR)

This 17-item test consisted of a combination of multiple-choice, completion, and

problem solving items. The items determined whether students had the knowledge needed

to participate in the simulation successfully. The test also determined whether paper and

pencil problems like those posed in the simulation could be corn pleted.

Transfer learning (TOL)

In this 10-item test problems were given to students that required skills similar to

those involved in the simulation. One problem required the sequencing of 6 events in a

logical order. Another involved moving furniture in a room in a systematic way to effect

195

a new arrangement. Five of the problems provided a grid system to students and required

them to either execute or design steps to move an object about on the grid. Three of the

problems presented a picture of a game board on which students were asked to plan moves

to gain high scores. The emphasis in these problems was on sequencing tasks to solve a

problem. In some respects the paper and pencil transfer problems were more difficult

than the computer simulation tasks because they did not show results after each step in

the problem, The student was required to envision the results of a series of steps or

possibly make some kind of drawing to show intermediate steps.

Tape recordings were made of selected student groups working with the computer

simultion in the computer laboratory. The purpose of these was to examine the tactics

and strategies students adopted and to determine how much conflict or cooperation

occurred in the groups. Individual interviews were also held with 6 students and tape

recorded. The interviews focused on understandings of the various tests the students

completed as well as their understandings of the simulation itself.

RESULTS

Descriptive statistics on all measures for the combined group and the three

individdal schools re given in Table 2.

Students working in pairs at the computers in School A had the most favourable

attitudes toward computing and the highest scores indicating the most knowledge of the

simulation (RAR quiz), but they still had the lowest scores on the transfer (TOL)

measure. Their logical thinking (GALT) scores were also lowest.

Students in School C that used the computer simulation the least had the least

favourable attitudes and knew the least about the simulation itself (RAR quiz) but still

scored highest on the transfer (TOL) measure, These students also had the highest logical

thinking scores of all students. Students from School B who worked in groups of four had

the intermediate scores on all measures.

Because the students differed in their logical thinking ability, a capability considered

important in successfully doing the simulation task, it was used as a covariate in an

analysis of the several outcomes used in the study. The results of this analysis are shown

in Table 2. Significant differences were found in attitudes toward computing and

knowledge of the simulation itself. In each case the two groups of students working with

the laboratory computers significantly outscored tile third group that used the computer

less and on an occasional basis. But the transfer of learning (TOL) scores show no

significant differences even though School C scores are substantially higher. Logical

thinking ability correlates strongly with transfer ability (r=.71) and the raw scores were

adjusted to bring these mean transfer scores closer together in the analysis of covariance,

196

TABLE 2

DESCRIPTIVE STATISTICS FOR ALL MEASURES

School A School B School C School Overall Max (Lab (Lab (Single Differen-

Score pairs) quartets) Computer -ces ** pairs)

n=50 n=15 n=l 5 n=20

Logical 3.24* I0 2.7 3.1 3.8 F= Thinking (1.7) (1.6) (2.0) (1.5)

Attitudes 80.0 96 82.8 82.7 75.9 F=7.79 Toward (7.5) (4.9) (6.6) (8.2) p=.001 Computing A B C

Raft Away 12.9 20 14.5 13.7 11.2 F=7.44 River Quiz (3.5) (4.0) (2.7) (2.9) p=.002

ABC

Transfer of 8.0 18 6.4 8.0 9.2 F=1.19 Learning (3.7) (3.9) (3. i) (3.7) p=.31

CBA

Each pair of values shows the mean score for the measure and the standard deviation of the scores in parentheses. Analysis of covariance results. Logical thinking ability is the covariate. Adjusted mean scores for the 3 schools are shown from left to right (highest to lowest) by the letters A, B, and C. Letters underlined by the same line are not significantly different from one another.

Analysis of the tape recordings revealed a variety of behaviours associated with

group work and strategies in carrying out the sire ulations.

* anticipating events "I'd better go get some fish before

or consequences the fire goes out."

* planning several moves "...then I'll wait for you to

ahead bring the wood up and I'll build

the raft."

* scenario of moves "...then we'll cut wood ... and

catch some fish ... and build

the raft ..."

* cooperative planning A:"Now you just keep cutting wood."

B:"OK"

C: "And I've got to go to the tree."

D: "And I've got to go to the

island."

Over the several class periods that students worked with the simulation it appeared

that cooperative behaviour increased. At the outset, students were more concerned with

197

their own turn than how their actions affected accomplishing the task. There were even

instances of ouright resistance to cooperation. One student, for example, wasn't going to

share any of the fish he had caught with others. As time went on students seemed to

cooperate more readily as it became clear that differentiation of roles and sharing of work

was essential to accomplishing the task.

One observation of students that stands out during the simulation was that they were

always on task - sometimes boisterously, occasionally argumentatively, and usually

excitedly, but always on task. Interest never flagged. Certainly there was novelty in use

of the computer although these students all had some prior experience with computers in

their schools. But the interactive, problem-orientated nature of the task completely

captured their attention. This occurred even though conditions in some cases seemed less

than ideal with 2 or 4 students competing for space in front of a single computer.

CONCLUSION

As students used the computer simulation more they clearly acquired more of the

knowledge needed and appeared to be able to solve paper and pencil problems like those

found in the simulation. Attitudes toward computing were also a bit more favourable for

students using the computer simulation the most. But these program gains did not

translate into greater ability to transfer skills to new, non-computer tasks. Students in

the group with the least knowledge concerning the simulation scored the highest on the

transfer test.

No formal measures of cooperative behaviour were used- only informal observations

of students as they worked together. It appeared that some of the initial individuality and

authoritarianism gave way to more cooperative behaviour as students realized how success

depended on working together.

To return to the point made early in this paper - a simulation such as Raft Away

River has little obvious value in and of itself. Schools would seem to not be interested in

whether youngsters can assemble small brown logs on a screen in the shape of a raft. This

is not a skill with any inherent value. But schools are concerned with effective outcomes

such as cooperative behaviour and with cognitive outcomes such as sequencing ability and

problem solving. To the degree that the Raft Away River simulation promotes these

outcomes, it has value.

Although transfer of learning appeared not to be strong in this study and not

differentially effective based on skill with the simulation, it shouldn't be concluded that

the computer program failed. First, the contact with the program was for only a few

hours so a modest amount of learning is all that can be expected. Problem solving ability

198

is a long term goal of schools contributed to by various subject areas over several years of

study. One short learning stint as in this study would be expected to influence problem

solving in only a small way. Secondly, measuring the ability to transfer or apply skills in a

new area is difficult because it is easy to misgauge transfer tasks. If the transfer is too

remote from the learning task, students are unable to see common elements in problems

and apply appropriate skills. If problems are set too close or similar to those in the

learning task, students don't have to comprehend a new setting and apply old skills to it.

Studies should be continued that seek evidence about the claims made for computer

learning programs. Reviews of computer programs by educators provide important points

of view about accuracy, completeness, and purpose. But they fail to give assurance of

whether using the computer software actually promotes the behaviours or objectives

specified by the developers. Such information would be valuable for all kinds of learning

materials but it is especially important for school activities that do not obviously promote

academic tasks.

REFERENCES

RAFT AWAY RIVER (1984). Jacaranda Wiley.

ROADRANGKA, V., YEANY, R. & PADILLA, M. (1983). The construction of a group assessment of logical thinking (GALT). Paper presented at the Annual Meeting of the National Association for Research in Science Teaching, Dallas, Texas.