A MULTI-USER EDUCATIONAL ONLINE GAME WITH WEB

BASED MATHEMATIC LEARNING BY USING ACTIVITY

THEORY ANALYSIS

Chi-Yuh Shen, Gwo-Dong Chen, Kuo-Liang Ou

Department of Computer Science and Information Engineering National Central University

No.300, Jhongda Rd., Jhongli City, Taoyuan County 32001, Taiwan(R.O.C.)

Department of Applied Mathematics

National Hisn-Chu University of Education

No. 521, Nan-Da Road, Hsin-Chu City, Taiwan, (R.O.C)

Keywords: Activity theory, intrinsic motivation, educational online game, machine learning, web based mathematic

learning.

Abstract: In this paper, we use activity theory to analyze the process and out come of mathematic learning by using

MEOG (Multi-User Educational Online Game) based on the previous research (Gwo-Dong Chen, Gee-Yu

Shen, 1997) described a multi-user educational online games which can motivate and guide students to learn

and practice courses on the web. In order to reduce the obstacle of MSEOG applying in instructional

activity in classroom, first, we adopt the factors of motivation as outcome and decided the tools and rule by

found the cue in the questionnaires and server log. Second, we consider events of instruction and learning

phrases (Gagne) and using Machine Learning Technique to get the potential rule (activity theory). From

these, we want to develop a methodology to get stability of guiding online game applying in instruction of

classroom.

1 INTRODUCTION

To play the game is the important part of our mental

and social development (Alan Amory, Kevin

Naicker, Jackie Vincent and Claudia Adams, 1998).

There are so many young men and younger adults

volunteer to spend much time to play online

game.(Alexamder E. Voiskounsky, Olga V. Mitina,

Anastasiya A.Avetiso, 2004). Recently, an online

game called “world of war craft” is popular in

Taiwan. It is a obvious evidence. Using game to

support in learning is not new research e.g.

increasing motivation (Westrom, Mary 1992, Gwo-

Dong Chen, Gee-Yu Shen, 1997,Maria M. Klawe

1999),enhancing achievement in mathematics

learning(Maria M. Klawe 1999), effect the cognitive

function (Rieber 1996) and training Skill to play

adventure games includes : logics, memory,

visualization, and problem solving (Quinn, 1994,

1997 Amory et al, 1999b).

In the study of Tom S. Chan (1999) mentions

”Deci and Ryan(1985) noted that making all school

work intrinsically motivating is impossible”.

Few research in game applied in school

activity.From BECTA surveying , Kirriemuir, JK,

McFarlane, A(2003) point out the main obstacle of

apply computer and video game in school is learning

in time, but teacher not allow student spend time in

the control of complex learning.

From E-GEMS project, Maria M. Klawe (1999)

got the result of gaming supporting learning are

effective increasing motivation and achievement.

But, in Muti-user network game, there is “significant

and persistent difficulties in bringing the games to

the level of stability needed to conduct field studies

of the games as a whole” (cited by Maria M. Klawe

1999).

In previous study (Gwo-Dong Chen, Gee-Yu

Shen, 1997), student volunteer to spend more time to

play MSEOG, besides, there are some factors of

motivation decided by T-test, but not study in

achievement and low rate in relearning the CAI.

312

Shen C., Chen G. and Ou K. (2006).

A MULTI-USER EDUCATIONAL ONLINE GAME WITH WEB BASED MATHEMATIC LEARNING BY USING ACTIVITY THEORY ANALYSIS.

In Proceedings of WEBIST 2006 - Second International Conference on Web Information Systems and Technologies - Society, e-Business and

e-Government / e-Learning, pages 312-319

DOI: 10.5220/0001252103120319

 SciTePress

A few study in using activity theory applied in

gaming related work. Activity theory provide a

method to analyzing learning processes and

outcomes (Lucia Rohrer-Murphy)

Kurt Squire 2002 argued the Game-Playing as

Social Practice and activity theory was proposed

under the issue-“how can one theoretical framework

for socio-cultural contexts” by psychologists. Victor

Kaptelinin Michael Cole 2002 adopted activity

theory in the research work about educational

computer game playing.Pippon Barr 2005 think

activity theory may prove a useful analytic approach

to the computer playing, and use activity theory to

develop a conceptual framework which describes the

role of value in computer game.

For this reason, this study based on the

advantage of combination online game and

instruction, issue of educational online game, using

activity theory, consideration events of instruction

and learning phrase (gagne) to develop the next

generation of educational online game applied in

instructional activity of classroom.

2 DEFINITION

Online game: players can interact in the same time.

Educational online game: combining CAI and

online game to achieve the goal of instruction.

Loosely mode educational online game: lower

learning support or intervene, player has the higher

degree of control in the game e.g. previous study

(Gwo-Dong Chen, Gee-Yu Shen, 1997). It let

student spend more time in learning. So, this model

could partial apply in classroom and continue

learning in out of school.

Tightly mode educational online game: higher

learning support or intervene, player has the lower

degree of control in the game e.g. this study. This

model designed to apply in instruction of classroom.

Mix mode educational online game: higher

learning support or intervene, player has the lower

degree of control in the game e.g. this study. This

model designed to apply in instruction of classroom

under tightly mode and adaptive tuning the mode in

out of school.

3 RELATED THEORY AND

RESEARCH

3.1 Intrinsic Motivation

The primary reason of embedding learning materials

into a multi-user computer game is to promote the

learning motivation of the students. Lepper and

Malone (1981) investigated what makes a computer

game fun. Malone,T.W.&Lepper, M.R.(1987)

found that motivation factors can be divided into

two groups: intrinsic motivation and extrinsic

motivation. The motivation factors of a single user

game are mostly related to intrinsic motivation. The

intrinsic motivation factors include:

Challenge: a game always provides missions to

be accomplished or virtual enemies to fight with. It

provides a challenge to the players;

Curiosity: a game always has a fantasy story with

graphic animation and sound. The missions provided

are like puzzles. A player is always trying to find

how to solve and what is in the fantasy scene;

Engagement: a game provides a simulated

environment. The players feel like they are the role

in the game for they always get response quickly

after they take actions. This keeps the players

focusing on the game. And, the players feel like they

are involved in the missions and environment;

Autonomy: a game provides a simulated

Tool

Rule

Division of labour

Sub

ects

Out Come

Communit

Figure 1: Yrjö Engeström’s activity system model.

A MULTI-USER EDUCATIONAL ONLINE GAME WITH WEB BASED MATHEMATIC LEARNING BY USING

ACTIVITY THEORY ANALYSIS

313

environment. Also, it provides actions the players

can take and control in order to achieve what they

expected. They do not have these capabilities in the

real world;

And Fantasy: a game can provide graphic

animation, music and sound, and actions with

response. These make a fantasy feel like real.

When Lepper and Malone (1981) studied

computer games, most computer games are only

single user games. Therefore, the factors considered

in what make computer games fun are related to

intrinsic motivation. However, in a multi-user game,

extrinsic motivation factors should be included. The

extrinsic includes

Competition: players in a multi-user game can

see the scores, level of other players. Besides, they

can compete or fight with other players;

Collaboration: a multi-user game can arrange

missions that are to be accomplished by group of

players. It can also provide an environment for

players to make friends;

And recognition: if a player got high score or

promote to higher level, other players can know

from the game. Thus, a multi-user game provides

chances for players to be recognized by other

people.

3.2 Activity Theory

Activity theory originates from Vygotsky’s idea

that explicit description of activity (Leont’ev 1978).

His student, Leont’ve, develops hierarchical levels

style from Vygotsky’s and Yrjö Engeström extended

Vygotsky’s to the activity theory (figure 1). Lucia

Rohrer-Murphy (1999) thinks that activity theory

provides a unique lens for analyzing learning

processes and outcome and activity theory is an

important precursor to good instructional

design.(figure 1)

3.3 Learning Stage (Gagne’s 1970)

The purpose of instruction is to assist learning

process (Robert M. Gagne, Leslie J.Briggs, and

Walter W. Wager 1988). The learning processes are

included in cognitive learning theory (Anderson,

1985; Estes, 1975; Klatzky 1980). They are

Attention, Expectancy, Retrieval of Relevant

Information to Working Memory, Selective

Perception, Encoding: Entry of Information into

Long-Term Storage ,Responding ,Feedback ,and

Cueing Retrieval.

3.4 Events of Instruction (Gagne’s

1970)

They deduce the event’s of instruction from learning

processes. They are Gaining attention, Activating

motivation: Informing the learner of the objective

,Stimulating recall of prerequisite learning,

Presenting stimulus material, Providing learning

guidance, Eliciting the performance, Providing

feedback ,Assessing the learner's performance,

Promoting retention and transfer.

3.5 Machine Learning

Technique-Decision tree

Decision tree: Decision Tree (Jiawei Han,

Micheline Kamber, 2001; Tom M. Mitchell, 1997) is

a tree structure similar to flow char structure. The

internal node represent the testing on attribute with

condition an branch represent the result of testing.

The leaf nodes represent the last result of

classification.

A typical example of Decision Tree (figure 2) is

that decision tree judge a customer whether buy the

car or not. The rectangle is equal to the node and the

ellipse is equal to the leaves. To judge one man

whether could buy the car or not, we have to get

some attributes of the man and use the attributes

tested by decision tree. The test is begun from the

root of the tree and test along to leaf. The leaf

represent the result of predict. For example: some

attributes of a man description below:

Table 1: Decision tree could easy transfer to these rules of

classifying (age=30, city=no, background=master).

Figure 2: Example of Decision tree.

If salary<=20 AND city=no

If salary<=20

AND city=yes

If salary =20…30

If salary =>30

AND background=master

If salary =>30

AND background=bachelor

THEN buys_car=no

THEN buys_car=yes

THEN buys_car=no

THEN buys_car=yes

WEBIST 2006 - E-LEARNING

314

Now, There are some decision support systems

of commercial use Decision tree to predict the

market of stock ,evaluate the risk of economic,.. and

so on.

3.6 A MUD Game

Many ideas of our multi-user game come from

Multi-User Dungeon MUD games. A MUD game is

a text base game. In a MUD game, there are (1)

levels of players and (2) communication tool such as

talk and chat room. Each level in MUD has different

privileges. Thus, it forms a players community.

Because it is a text base game, it provides users

imagination possibilities. However, it is not suitable

for elementary school students. We adopt its ideas to

form a community of practice.

3.7 Study of Westrom and Shaban

Westrom and Shaban (1992) investigated the effect

difference between a game and an instructional

game. They showed an instructional game is not so

fun as a pure game. In this paper, we investigate the

degree of effect for motivation factors proposed by

Lepper and Malone. We compare the difference

between a pure tutorial course and an instructinal

game.

Figure 3: Arranging the game script according to the learning hierarchy.

Consecutive

uZa county

subtraction

Consecutive zon-

han

multi

lication count

Consecutive Na-kan counity

addition

Mix of Chen-Der county

addition and su

traction

Prepare

ain scene

Using parenthesis Da-

n county

Consecutive Da-Ton

division count

Mix of yan

-ming cave

consecutive multiplication,

consecutive division and parenthesis

Mix of addition, su

traction, Da-Tuen cave

multiplication, division, and

parenthesis

Token A and, score>=200

oken D and score>=2000

Token E and score>=3000

Token F, G and score>=4000

Token H and score

>=5000

Token B, C and score >=1000

A MULTI-USER EDUCATIONAL ONLINE GAME WITH WEB BASED MATHEMATIC LEARNING BY USING

ACTIVITY THEORY ANALYSIS

315

3.8 WebQuest

Web Quest (CoDa96) is a web based instructional

game. The students ramble about in the game to find

related information to pass a scene. When facing

with problems, the students can invoke a web

browser to find the answer. The students can design

their own game. Experiment shows that the students

more like to design a game than to play the game.

They did not try to build a community of practice in

the game.

4 GAME SCRIPT AND

LEARNING MATERIALS

4.1 Selection of the Style of the

Game

The first thing of developing a instructional game is

to design a game script and put learning materials

into the game. Since a learning material always has a

learning hierarchy (Gagn68), the game script should

follows the learning hierarchy. Therefore, we decide

to implement our game as a role playing for the

following reasons.

First, role playing games are the most popular

games in Taiwan and this style of games allows

player to explore and think. A player is required to

figure out the structure of the game and learn the

strategies to overcome the problems in order to

accomplish the missions given in the game script

and to get score.

Second, a role playing game always has a

hierarchical script structure. Thus, it is easy to

design the game script according to the learning

hierarchy. When the learners try to accomplish the

script’s mission, they will learn and practice the

learning material according to the learning

hierarchy.

Third, a multi-user role playing game can forms

an environment to serve as a community of practice.

We can design different roles in the game. Each role

has different privileges. A role can invoke an action

only if it has the privilege to take the action. Thus,

the game scenery forms a community of practice if

every role is using the learning topics to solve

problems in order tp accomplish missions. However,

there should be missions assigned for roles to

cooperate or collaborate.

4.2 Arranging the Game Script

According To the Learning

Hierarchy

Gagne (Gagn68) proposed that a learning material

can be decomposed into elements and learning

should follow a procedure from simple to complex.

Thus, learning material should organized as a

learning hierarchy with a learning flow embedded in

it. This structure can be perfectly fit into a role

ule

Rule of game

Learning condition

Division of labour

(Role)

Master, sheik, and

ordinary people

Sub

ects

Player / NPC

Out Come

(To spend more time in

learning and **Curiosity,

**Challenge,

*Competition,*Cooperation,

*Cognition)

Communit

TOOL

MSEOG ( **querying and displaying of NPC and **learning

companion, **help me,**teaching, **questions, **chat room, **peer

talk, **note book, *represent of role,*shout), pen, paper, calculator

of computer, CAI

Figure 4: MSEOG based activity theory.

WEBIST 2006 - E-LEARNING

316

playing game script. Figure-1 is a learning hierarchy

for arithmetic skills. Each box represents a subskill.

The arrow line represent the legal learning flow. The

bar line represents a and join. A student should learn

all the subskills connected to the bar line.(figure 3)

We can design a hierarchy of script scenery

based on the learning hierarchy as shown in Figure-

1. The text in italic form are of the game script. Each

learning element or box is a scene or an act. In the

scene, the background is a place in Taipei city with

roles in it which give hints, asking question, and

lecturing through page in world wide web. The

player is trying to get scores and the token by

studying courseware or practicing exercises.

Players should get enough score or token in order

to get into another act or scene. The condition of

passing is stored in the bar line in the above figure.

The players are free to go to any scene if they hold

enough tokens and scores. Since the game script is

designed based on the learning hierarchy and flows,

the players will learn and practice accordingly.

5 ACTIVITY THEORY

ANALYSIS-ONLY CAI AND

MSEOG WITH CAI IN

INSTRUCTION OF

CLASSROOM

We used the factor of motivation of learning and

server log to serve as evidences to analyze activity

of MSEOG in previous study. We found some

questions of questionnaire related to the factors and

decided some tools involved in the activity from the

cue in the questions. We choose the most high score

one of questions. They are listed below then tools

and rules decided by the cue of question:

Curiosity: I want to know another player when I

played the MSEOG. From this question, querying

and displaying of the NPC and learning

companion were the tools to do that.

Challenge: I continue to play next stage of game

when I passed the condition of learning. From this

question, question in the game is the tool. The

challenge comes from two ways. One is the NPC

and the other is learning companion. A player need

help when answers the question. He could use “help

me” function by clicking animation of learning

companion. The challenge is to teach. In order to

teach, we provide discussion and synchronize

action function to reach “let me do it show you”.

Discussion function includes peer talk, chat room.

A player could challenge another play via

“challenge you” function with notebook to choice

question when another higher degree role player use

the “shout” function to show this position in the

learning community designed by considering the

Cognition of motivation (ML87).

Competition: I feel very happy when I passed the

condition of learning. From this question, condition

of learning is the rule.

Cooperation: I want to discuss with classmates

after playing game and discuss seriously for

answering question in the future to reach the

condition of learning to go next stage of game. From

this question, language, paper and pen are the

tools.

Cognition: I feel exciting when I got the higher

degree of role. From this question, the role

representation and “shout” function are the tools.

To support the decision, we analyze the server

log. We found the coincidence between curiosity to

comparison for self and another and challenge to

question answering by T-test significant degree and

percent order.

In the server log, the communication by text

supported the peer talk and chat room tool using and

teach mode support the synchronize action.

The purpose of the previous study is focus on

motivation supported by game, so the CAI

presentation served as the same environment. But

there is a problem generated from previous study

about CAI presentation, e.g. how to identify the

meaningful learning of frequency of CAI presenting.

Play Action Count Percent (%)

Communication by Text 558 14.11

Teach mode 61 1.54

Ask question to another 100 2.53

Choice style question

answer

1037 26.24

Comparison for self &

another

1401 35.45

CAI presentation 36 0.91

Virtual people hint 759 19.21

6 GUIDLINE OF APPLING

EDUCATION ONLINE GAME

IN CLASSROOM

By result, there are some guidelines to apply

education online game in classroom in first stage by

considering learner feeling.

A MULTI-USER EDUCATIONAL ONLINE GAME WITH WEB BASED MATHEMATIC LEARNING BY USING

ACTIVITY THEORY ANALYSIS

317

1. “Learning condition” is the most important

rule of game. It mediates the subjects in the

community to use tool to let mathematic learning

transform to outcome-Competition increasing.

2. “Querying and displaying of NPC and learning

companion” are the most important tools to let

subjects to transform mathematic learning to

outcome-curiosity enhancing.

3. “Learning companion, help me, teaching,

questions, chat room, peer talk, note book” are the

same as important tools to let subjects to transform

mathematic learning to outcome- Challenge.

4. “Represent of role, shout”- are the second

important tools to let subjects to transform

mathematic learning to outcome- Challenge

enhancing.

5.”shout” is the same as the second important

tool to let subjects to transform mathematic learning

to outcome- Cognition enhancing.

6. Above 5 points based on students’ feeling

about intra-game situation, we also found student

paper, pen, and calculator in PC to assist the solving

question by using observation based on extra-game

situation.

7 POTENTIAL RULES

GENERATION BY MACHINE

LEARNING TECHNIQUE

There are two kinds about generating potential:

1. one activity :Using learning action of Server log

to predict motivation/achievement, for example:

(simulation data)

34 question answering > 72:

: 11 CAI > 47:

: :...14 peer to talk <= 26: 49 (7/5)

: 14 peer to talk > 26:

: :...29 challenge you <= 28:

: :...10 teach mode <= 36:

: : :...35 notebook <= 60: 78 (3/1)

: : : 35 notebook > 60: 40 (2)

: : 10 teach mode > 36:

: : :...4 decide to PK <= 65: 67 (2)

: : 4 decide to PK > 65: 62 (2)

: 29 challenge you > 28:

: :...18 roam(thinking or idle) <= 33: 92 (4/2)

: 18 roam(thinking or idle) > 33:

34 question answering <= 72:

:...5help me > 58:

:...25shout <= 24:

: :...15 express emotion <= 38:

: : :...15 peer to talk <= 16: 89 (3/1)

: : : 15 peer to talk > 16: 44 (2/1)

: : 15xpress emotion > 38:

: : :...6 notebook <= 52: 24 (3/2)

: : 6 notebook > 52: 11 (3/2)

: 25shout > 24:

: :...9 peer to talk > 63:

: :...21 express emotion <= 45:

: : :...34 let me do it show you <= 46: 63 (3/1)

From this perspective, we found a problem in

using server. The frequency of learning action was

not explained the learning meaning related to

achievement. For explain: peep to talk so many

times why achievement is higher? So, we have to so

some strategy to prove the learning action meaning

quality in learning.

2. between two activities: Using factors of

motivations to predict achievement.

8 RESEARCH LIMITATION

There are two limitations in previous research.

First, because it is difficulty of measuring

motivation, using the same questionnaire in only

CAI and online game with CAI is the limitation.

Second, chidden whether the ability of expression

the motivation has or not. Finally, it is difficulty to

find the relation between with intrinsic motivation

and learning achievement in first stage of education

online game learning.

9 CONCLUSTION

This study based on previous study to analyze the

Multi-user educational online game applying in

instruction of classroom. the motivation as a

outcome, we backward to find the cues in the

questions of questionnaire. Then, we also analyze

the server log. There are some decision about

activity theory: learning condition is a potential rule.

MSEOG ( querying and displaying of NPC and

learning companion, questions, chat room, peer talk,

note book, represent of role), pen, paper, calculator

of computer, CAI are tools.

In order to reduce the obstacle of MSEOG

applying in instructional activity in classroom, first,

we adopt the factors of motivation as outcome and

decided the tools and rule by found the cue in the

questionnaires and server log. Second, we consider

events of instruction and learning phrases (Gagne)

and using Machine Learning Technique to get the

WEBIST 2006 - E-LEARNING

318

potential rule (activity theory). From these, we want

to develop a methodology to get stability of guiding

online game applying in instruction of classroom.

There are six guide found in the first stage

describing before.

Finally, there are two advantages of constructing

the education online game community practice. One

is the teacher has opportunity to conduct the social

learning in the gaming environment so as to avoid

the frustration or incorrect action in social

interaction. Other is the environment will extend the

learning in classroom to out of school and to

facilitate student continue learning about knowledge

in classroom.

REFERENCES

Amory, A., Naicker, K., Vincent, J. & Claudia, A. (1998).

Computer Games as a Learning Resource.

Proceedings of ED-MEDIA, ED-TELECOM 98,

World Conference on Education Multimedia and

Educational Telecommunications, Vol. 1, pp. 50-55.

Alexander E. Voiskounsky, Olga V. Mitina, Anastasiya A.

Avetisova Playing Online Games: Flow Experience

PsychNology Journal, 2004 Volume 2, Number 3, 259

- 281

Chan, T. S., & Ahern, T. C. (1999). Targeting Motivation

- Adapting Flow Theory toInstructional Design.

Journal of Educational Computing Research, 21(2),

151-163.

Gwo-Dong Chen, Gee Yu Shen, Yuan-Jen Tai & Alf Lai

(1997) A Multi-User Situated Learning Games on

Internet ICCE.

Gagné, R., Briggs, L., & Wager, W. W. (1988). Principles

of instructional design (3rd ed.). New York, NY: Holt,

Rinehart and Winston, Inc.

Jiawei Han, Micheline Kamber, Data Mining : Concepts

and Techniques, MORGAN KAUFMANN

PUBLISHERS , 2001

Kirriemuir, JK, McFarlane, A, (2003) Use of Computer

and Video Games in the Classroom. Proceedings of

the Level Up Digital Games Research Conference,

Universiteit Utrecht, Netherlands. Available from:

http://www.silversprite.com/

LEONT’EV, A. N. Activity, Consciousness, and

Personality. Prentice Hall, 1978. Available online at

http://www.marxists.org/archive/leontev/works/1978/.

Lucia Rohere-Murphy (1999) Activity theory

LAWRENCE ERLBAUM ASSOCIATES, Publishers

Mahwah, New Jersey p.159

Maria M. Klawe: Computer Games, Education and

Interfaces: The E-GEMS Project. Graphics Interface

1999: 36-39

Malone,T.W.&Lepper, M.R.(1987) ”Making Learning

Fun: A Taxonomy of Intrinsic Motivations for

Learning,” In R.Snow and M.Farr (Eds.)﹐ Aptitude,

learning, and instruction ﹐ vol.3 ﹐ Cognitive and

affective process analysis﹐ Hillsdale, NJ: Lawrence

Erlbaum Assoc.﹐ 1987﹐ ch.10﹐ pp.225-254

Malone, T.W. & Lepper, M.R. (1987).”What makes

computer games fun,” Byte,6(12),258-277.

Pippin Barr 2005 Play and Work:Investigating the Role of

Value in Human-Computer Interaction A proposal

submitted to the Victoria University of Wellington

RM Gagne70 The conditions of learning, New York, Holt,

Rinehart and Winston

Squire, K. (2002). Cultural framing of computer/video

games. The International Journal of Computer Game

Research, 2(1), Game Studies Available

online:http://www.gamestudies.org/0102/squire/

Retrieved 12-24-02

Rieber, L.P. (1996). Seriously considering play: Designing

interactive learningenvironments based on the

blending of microworlds, simulations, and games.

Educational Technology, Research and Development,

44, 43-58.

Westrom M. and Shaban A. （ 1992 ） “Intrinsic

Motivation in Microcomputer Games”Journal of

Research on Computing in Education﹐ vol.24

﹐ no.3

Summer 1992﹐ pp.433-445﹒

Quinn, C.N. (1994). Designing educational computer

games. In K. Beattie, C.McNaught, & S. Wills, (Eds.)

Interactive multimedia in university

education:Designing for change in teaching and

learning, pp. 45-57, Amsterdam:Elsevier.

A MULTI-USER EDUCATIONAL ONLINE GAME WITH WEB BASED MATHEMATIC LEARNING BY USING

ACTIVITY THEORY ANALYSIS

319