Original article

Internet project-based learning environment: the

effects of thinking styles on learning transfer
C.-I. Lee & F.-Y. Tsai
Institute of Computer Science & Information Education, National Tainan Teachers College, Tainan, Taiwan

Abstract The purpose of this study, in an environment of Internet project-based learning, is to

undertake research on the effects of thinking styles on learning transfer. In this study, we
establish an environment that incorporates project-based learning and Internet. Within this
environment, we divide our sample of elementary school students into four groups:
Executive Group, Legislative Group, Judicial Group, and Mixed Group. Taking the learning
of ‘Natural Science’ as an example, we investigate the effects of different thinking styles on
learning transfer. The results of this study are:
(a) significant differences between the near transfer of the Executive Group and the Legi-
slative Group,
(b) no significant differences in far transfer are found among groups of different thinking
styles,
(c) the near transfer of the Mixed Group is superior to that of the Legislative Group and the
Judicial Group, and
(d) the far transfer of the Mixed Group is superior to that of the Legislative Group.

Keywords cooperative learning, elementary school student, Internet project-based learning, perfor-
mance assessment, thinking styles, transfer of learning

chers’ guidance, enables students to explore and solve

Introduction
problems together with their peers. Project-based
With the great progress of science and technology, the learning has the same abbreviation ‘PBL’ as problem-
information industry has grown enormously, which in based learning, but is somewhat different. The former
turn has caused fundamental changes in people’s adopts a production model. Its whole process is
learning styles and has helped usher in a knowledge- authentic and reflects the real production activities.
based economy and more skilled division of labour. In The tasks on hand will be finished using the ideas and
the meantime, we urgently need to teach students at ways that students cherish. Although the artefact is the
a faster pace or guide them in ‘learning how to learn’ driving force of PBL, the most important essence to
and in ‘cooperative learning with others’. Project- succeed learning is how to acquire contents, knowl-
based learning (PBL) places the most emphasis on edge, and the skills from the process of production. In
topics of everyday concern, which, through the tea- comparison, problem-based learning adopts an inquiry
model, which regards one problem as its driving force,
Accepted: 15 August 2003 but may concentrate on answering to some extent.
Correspondence: Chien-I Lee, Institute of Computer Science &
Some types of problem-based learning require stu-
Information Education, National Tainan Teachers College, 33, Sec. 2,
Su-Lin St., Tainan 700, Taiwan. dents to define questions definitely, set up hypotheses,
Emails: leeci@ipx.ntntc.edu.tw, tsai.funyuan@msa.hinet.net collect materials, and describe answers clearly. Some

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39 31
32 C.-I. Lee & F.-Y. Tsai

types of problem-based learning are looked on as learned, including knowledge and skills, in another
learning-embedded cases, in which there is no definite situation (Broad 1997; Perkins & Salomon 1988). The
answer; they are only used to attract students to study transfer of learning can be classified into positive
and collect materials. Students may or may not deliver transfer and negative transfer, depending on its nature.
end production and teachers will grant them enough Positive transfer can be further classified into near
time for self-questioning and self-assessment. transfer and far transfer according to the differences of
PBL has proved to be an effective teaching and degree. Near transfer means solving problems with
learning strategy (Marchaim 2001; Murphy & Yakut similar characteristics or carrying out some learned
2001; Huang et al. 2002). This approach to teaching skills in real contexts resembling learning environ-
and learning has been gaining more and more favour ments. For example, learning the addition of digital
recently. The main reason is that it is in accordance numbers helps to learn the addition of trinary numbers.
with the learning concepts of constructionism, cogni- Far transfer means the ability of solving different
tive psychology, and integrated curriculum. Further- problems in different learning environments; i.e. the
more, it is an instructing-learning pattern that learner has to possess the ability to extract the com-
combines curricula, teaching methods, and assessment mon principles derived from one kind of learning
into one single unit. However, in classroom teaching context and use them in all-new learning contexts. For
and learning contexts in traditional schools, because of instance, learning the addition of digital numbers
the factors of limited classroom space, teaching time, helps in learning multiplication. Therefore, we know
etc., the planning of cooperative learning is a difficult that the transfer of learning has always been an im-
job, which is almost impossible to implement thor- portant goal in learning and teaching, and has drawn
oughly in schools (Marx et al. 1997; Balakrishnan much attention from the authority concerned. How-
2000). Thus, if we want the students to undertake ever, it is not an easy goal to reach (Davis 1999).
PBL, it might not work without new technologies to Among the factors relating to success or failure, the
support this new learning process (Laffey et al. 1998). questions of whether the teaching environment can be
With a view to this, how to use technology advances in learner-centred and bring forth the various issues
multimedia computers and the Internet to make relating to the students’ life are very essential
NetPBL (Internet project-based learning) a useful (Baldwin & Ford 1988; Macaulay 2000). Fortunately,
platform and enable learners to overcome the limits of PBL possesses these necessary aspects of teaching and
time and space as well as take part in versatile dis- learning.
cussions is an urgent question. When we answer it we Furthermore, from Sternberg’s theory of mental
should be able to broaden the scope for interaction self-government, we know that thinking styles play an
between learners and others. In short, this is an important role in learning (Sternberg 1997; Francisco
important topic for research. This phenomenon can be & Elaine 2000). The expression ‘thinking styles’ re-
shown by two trends: (a) the employment of PBL in fers to personal preferences in employing one’s in-
research and learning in all fields are on the rise telligence and competence when thinking or dealing
(Laffey et al. 1998; Balakrishnan 2000; Marchaim with things. Thus, two persons of equal ability might
2001); and (b) most researchers agree with the show totally different levels of performance (Stern-
teaching and learning strategy of PBL, and think that berg 1988, 1994). Basically, different people have
using the Internet as a basic instrument is the wave of different styles, which have nothing to do with being
the future (Barry et al. 2001). good or bad. The question is whether we can fit the
Although NetPBL has become a modern learning styles to the demands of the environment, for students
trend, whether it can accomplish its educational target, to fully develop what they are good at. This theory
i.e. the transfer of learning, remains a question that compares mental functioning to a political body,
merits study. The transfer of learning means the phe- dividing the self-management area of individual
nomenon of expanding or generalizing learning results. mentality into five dimensions, which in turn consist
It is an influence of one kind of learning on another. of 13 thinking styles. Among them, the division into
Some researchers have suggested that its purpose is to legislative, executive, and judicial styles by their
instruct students on how to utilize what they have functions are by far the most striking features. The

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39
Internet project-based learning environment 33

Legislation Style is concerned with re-creating,
In the NetPBL environment, the near transfer of
imagining, devising, and planning; the Executive Style the Mixed Group is superior to that of the single-
has to carry out actions decided upon by the Legisla- thinking-style group.
tive; and the Judicature is involved in judging, eval-
In the NetPBL environment, the near transfer of
uating, and comparing, so as to supervise the Mixed Group is superior to that of the Executive
performance of the above two styles. Judging from the Group.
above, we can deduce another issue worth studying;
In the NetPBL environment, the near transfer of
i.e. the working together of persons possessing different the Mixed Group is superior to that of the Legis-
thinking styles may bring forth better cooperative lative Group.
results, just as the case of government, which needs
In the NetPBL environment, the near transfer of
different mechanisms of execution, legislature, and the Mixed Group is superior to that of the Judicial
judicature to produce the greatest efficiency. Group.
Therefore, to conclude the above discussion on
In the NetPBL environment, the far transfer of the
NetPBL, transfer of learning, and thinking styles, we Mixed Group is superior to that of the single-thinking-
note that the previous researches were mainly con- style group.
ducted using traditional methods. However, the dis-
In the NetPBL environment, the far transfer of
crepancy between traditional and Internet learning the Mixed Group is superior to that of the Executive
contexts is great. Under traditional learning contexts, Group.
research of thinking styles cannot be analogized to the
In the NetPBL environment, the far transfer of
Internet learning environment. Furthermore, we know the Mixed Group is superior to that of the Legis-
from the theory of mental self-government that the lative Group.
possibility of bringing together people with different
In the NetPBL environment, the far transfer of
thinking styles having better cooperative results is the Mixed Group is superior to that of the Judicial
merely a theoretical assumption, still awaiting ver- Group.
ification. What is more, we cannot find any discussion
of the influence of thinking styles on transfer of
learning in the relevant literatures. Thus, the purpose Method
of this study is, taking elementary school students’ Participants
learning about ‘Natural Science’ as an example, ac-
cording to the grouping of thinking styles, to explore This study used fifth-year students as experiment
the influence of different combinations of thinking samples (n 5 156). The students’ thinking styles
styles on the students’ capacity for learning transfer in served as the independent variable. They were divided
the NetPBL context. into Executive Group, Legislative Group, Judicial
Group, and Mixed Group; the Mixed Group consisted
of the thinking styles of the former three groups. They
Hypotheses of the study all did their learning in the NetPBL environment.
Based on the previous description of PBL, learning Every student was randomly assigned to one of the
transfer, and thinking styles, which contribute to stu- above-mentioned four groups. Each group had three
dents’ academic achievement (Grigorenko & Stern- students, and there were 52 groups altogether.
berg 1997; Sternberg 1997; Zhang & Sternberg
1998; Zhang, 2002a, b), we propose the following
A NetPBL environment
hypotheses.

In the NetPBL environment, there are significant This study attempted to establish an environment that
differences in near transfer among groups of different incorporates the features of PBL and Internet, whose
thinking styles. system structures can be seen in Fig. 1. Figure 2 depicts

In the NetPBL environment, there are significant the homepage of the website (http://digger.ptnd.tn.
differences in far transfer among groups of different edu.tw/). This system is primarily divided into the
thinking styles. students’ learning interface and the teachers’ teaching

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39
34 C.-I. Lee & F.-Y. Tsai

Other functions: There are some handy functions

that are very convenient for students. For example,
Teacher Students
‘Calling teacher’ is a function that students can use to
call their teachers and consult them at any time during
the discussions, and ‘Calling fellow students’ provides
Managing project Choosing class and project
students with tools to have an immediate discussion
Developing subquestions with intra- or inter-group members.
Observing and monitoring of
learning process Learning benchmark lessons

Editing benchmark lessons Disscussing and investigating Teachers’ teaching interface

Editing schedules Developing calendars
Teachers’ managing functions: Here teachers can
observe all groups when developing the process of
Managing files Arranging artifacts
‘subquestions’, ‘investigation’, etc. This can be used
Assessing artifacts Communicating real time
to help students solve their problems or participate in
the discussion, allowing teachers to become assistants
and co-learners of the projects. In addition, teachers
Database of can also edit benchmark lessons, overlook or assess
learning process
Record
finished group artefacts, and monitor all kinds of
files of administrative affairs, such as registration and group-
various ing of students online.
functions
Learning process recorder: This recorder stores all
the students’ data concerning the learning process in
Database NetPBL. In future, teachers can use it to carry out
research and analysis in order to understand the stu-
dents’ learning patterns when using certain advanced
Fig. 1 System frame of NetPBL.
database technology (e.g. data mining) (Han & Kam-
ber, 2001). By this means, teachers can assess the
interface. The following points provide a brief in- students’ learning results and make plans for sub-
troduction of the features and functions of this system. sidiary teaching.

Procedure and methods

Students’ learning interface
Developing the subquestions: After the teacher de- This experimental teaching lasts for 3 months. To
cides on the ‘driving question’ and explains it to the ensure a clear explanation of it, we have divided it into
students, the latter can ponder on ‘subquestions’ three parts: preparatory activities, teaching and
arising from the driving question. Meanwhile, the learning activities, and assessment.
students can also carry on synchronous and asyn-
chronous discussion with the other members. By so Preparatory activities
doing, they can reach an agreement by themselves on (a) Project-conceiving: Researchers and teachers
how to produce subquestions. came to an agreement that they would use the unit
Investigations: In this section, students can discuss ‘The speed and directions of moving objects’ in the
issues concerning their projects with their group ‘Natural Science’ course, studied by fifth graders in
members. Of course, they can also post online what elementary school, as the teaching and learning con-
they have discussed with other members, or mention tent, in which ‘What is energy?’ would serve as the
their own original ideas. driving question. This accords with six principles:
Arranging and sharing artefacts: Students can feasibility, worthiness, contextualization, meaning-
produce their own homepages online, which is a fulness, sustainability, and ethics. These are suitable
showroom for each group to arrange and share their for use as driving questions (Krajcik et al. 1999).
artefacts. (b) The building of the NetPBL learning system.

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39
Internet project-based learning environment 35

Calling Teacher Developing Subquestions Developing Calendars Benchmark Lessons

Calling Fellow Students Arranging Artifacts

Investigations
General Tools Group Tools Keeping a Diary

Fig. 2 Homepage of NetPBL.

(c) The testing of Thinking Styles Scale and group- members by using the interface of ‘developing sub-
ing: Based on the results from the Thinking Styles questions’; they also had to share their findings and
Scale of Elementary School Students, we assigned the express their own ideas by making use of the interface
students randomly to each group. The scale that we of ‘investigations’; after that they had to observe the
used in this study was devised by Lin (1999) and timetable planned by the interface of ‘Developing
adapted from the Sternberg–Wagner Thinking Styles Calendars’ in order that they could take control of the
Inventory, which was compiled by Sternberg and learning process of each group.
Wagner in 1991. Lin’s adaptation has added the (d) Arranging artefacts: Each group manifested and
learning experiences of elementary students and the shared the results of their work and process to the class
teaching and learning contexts of elementary schools by using the interface.
in Taiwan into the methodology. It is suitable for use
from fourth to sixth graders in elementary school. Assessment
(d) Editing benchmark lessons and every rating scale This study assessed the learning transfer of the stu-
needed to prepare for performance assessment. dents by means of performance assessment, which
places emphasis on turning knowledge and under-
Teaching and learning activities standing into practical action, so that the teachers can
(a) The teacher explained to the students learning judge their students’ learning results by means of their
objects, learning procedures, and the presentation of planning, organization, and use of applied knowledge
finished artefacts of NetPBL. and skills. Evidently, performance assessment evalu-
(b) The teacher instructed benchmark lessons to stu- ates not only what knowledge students possess but
dents to supply them with preparatory knowledge for also how they use this knowledge to solve problems
investigating their projects. (Brown & Shavelson 1996). Thus, it is a good way to
(c) Investigations: After gaining knowledge and skills assess transfer of learning. The performance assess-
from benchmark lessons, students were able, under the ment scales include the ‘Learning Transfer Pretest
guidance of the teachers, to work with their group Assessment Scale’, the ‘Near Transfer Assessment
members to investigate and research their projects, in Scale’, and the ‘Far Transfer Assessment Scale’. For
class or after class. During this time, they were able to complete learning transfer scales, we have ‘Manual
discuss will-be-going ‘subquestions’ with their group and Assessment Criteria for Teachers’, ‘Performance

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39
36 C.-I. Lee & F.-Y. Tsai

Assessment Test Sheets for Students’, ‘Assessment Table 2. Summary of ANCOVA: the near transfer of different-
Notebook’, and ‘Checklist’, etc. thinking-styles groups

Sources of variation SS 0 df MS 0 F o2
Data analysis
Between groups 106.03 3 35.34 9.71
The data collected by this study included scores (thinking styles)
achieved by students in ‘Natural Science’ and ‘Com- Within groups 171.13 47 3.64 0.339
(error term)
puter Science’, the ‘Near Transfer Assessment Scale’,
Total 277.16 50
and the ‘Far Transfer Assessment Scale’. The statistic
method used was the one-way ANCOVA. We used the ANCOVA 5 analysis of covariance, df 5 degree of freedom.
Po0.01.
thinking styles of students as independent variables,
the learned near transfer and far transfer as dependent Table 3. Summary of analysis: a posteriori comparisons of
variables, and the scores achieved in ‘Natural Science’ Table 2 using Scheffe’s method
and ‘Computer Science’ as covariates. When the sta-
Thinking styles Mean variances F
tistic results were found to be significant, we conducted
of two groups
a posteriori comparisons using Scheffe’s method. (I) (J)

Exe Leg 3.16 18.16

Results Exe Jud 1.51 3.68n.s.
Near transfer Exe Mix 0.96 1.33n.s.
Leg Jud 1.65 3.94n.s.
From Table 1 we observe that the Mixed Group gained Leg Mix 4.12 21.21
the highest mean scores in near transfer, followed by Jud Mix 2.47 11.09
Executive Group, Judicial Group, and Legislative Exe 5 Executive Group, Leg 5 Legislative Group, Jud 5 Judicial
Group. Even when the mean scores of ‘Natural Sci- Group, Mix 5 Mixed Group, n.s. 5 not significant.
Po0.05; Po0.01.
ence’ and ‘Computer Science’ of these four groups were
adjusted to the same levels, the sequences remained
the same. From Table 2 we observe further that, after method to conduct a posteriori comparisons, which
removing the influence of the covariance, F 5 9.71 are summarized in Table 3.
reached a level of significance of 0.01. This means that From Table 3 we observed that the treatments of the
in the NetPBL environment, there were significant first and second, second and fourth, and third and
discrepancies in the near transfer among the thinking fourth experiments, after using Scheffe’s method,
styles of different groups. If we take a closer look at resulted in F values of 18.16, 21.21, and 11.09 re-
the strength of association (o2 5 0.339), we can see spectively, which were higher than the critical value
that thinking styles and near transfer correlate to each F 0 5 12.75 and 8.43, meaning that the level of sig-
other, and that thinking styles can account for 33.9% nificance of 0.01 and 0.05 were reached. That is, the
of the total variation of near transfer. In order to arrive Executive Group and the Legislative Group, the
at a better understanding, we employed Scheffe’s Legislative Group and the Mixed Group, and the
Judicial Group and the Mixed Group manifested a
Table 1. Summary of means, SD, and adjusted means: the near level of significance in near transfer.
transfer of different thinking styles groups

Groups No. of groups Means SD Adjusted means Far transfer

Exe 13 9.79 2.17 9.50 Table 4 reveals that in far transfer, the Mixed Group
Leg 13 6.86 2.29 6.34 gained the highest mean scores, followed by Executive
Jud 13 7.71 1.61 7.99 Group, Judicial Group, and Legislative Group. After
Mix 13 9.93 1.74 10.46 adjusting the mean scores of ‘Natural Science’ and
Exe 5 Executive Group, Leg 5 Legislative Group, Jud 5 Judicial ‘Computer Science’ of these four groups to the same
Group, Mix 5 Mixed Group. levels, we still obtained the same sequencing. Further-

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39
Internet project-based learning environment 37

Table 4. Summary of means, SD, and adjusted means: far Table 6. Summary of analysis: a posteriori comparisons of
transfer of different-thinking-styles groups Table 5 using Scheffe’s method

Groups No. of groups Means SD Adjusted means Thinking styles Mean variances F
of two groups
Exe 13 8.27 1.77 8.1 (I) (J)
Leg 13 6.55 2.09 6.24
Jud 13 7.19 1.57 7.36 Exe Leg 1.86 7.12n.s.
Mix 13 8.86 1.37 9.17 Exe Jud 0.74 1.00n.s.
Exe Mix 1.01 1.87n.s.
Exe 5 Executive Group, Leg 5 Legislative Group, Jud 5 Judicial Leg Jud 1.12 2.05n.s.
Group, Mix 5 Mixed Group. Leg Mix 2.93 10.65
Jud Mix 1.81 6.83n.s.
Table 5. Summary of ANCOVA: far transfer of different-thinking-
Exe 5 Executive Group, Leg 5 Legislative Group, Jud 5 Judicial
styles groups
Group, Mix 5 Mixed Group, n.s. 5 not significant.
Po 0.05.
Sources of variation SS 0 df MS 0 F o2

Between groups 47.07 3 15.69 4.89
In the NetPBL environment, the near transfer of
(thinking styles) the Mixed Group is superior to that of the Legislative
Within groups 150.97 47 3.21 0.186
Group and the Judicial Group.
(error term)
Total 198.04 50

In the NetPBL environment, the far transfer of the
Mixed Group is superior to that of the Legislative
ANCOVA 5 analysis of covariance, df 5 degree of freedom. Group.
Po0.01.
There are certain similarities in the results of this study
and other relevant studies as well as differences be-
more, Table 5 shows that after removing the influence tween them. For example, Sternberg (1997) contends
of covariance, F was 4.89, reaching a level of sig- that consideration of different thinking styles can help
nificance of 0.01. This means that in the NetPBL develop the best teaching and learning methods.
environment, the far transfers of different-thinking- According to his theory, NetPBL is a cooperative
styles-groups are significantly different. If we take a group learning method, in which students can discuss
deeper look at strength of association (o2 5 0.186), we their own projects freely and plan their own project
can see that although there is a correlation between schedules as well as investigation procedures. Judging
thinking styles and far transfer, thinking styles can only from the research results of Sternberg (1997), the
explain 18.6% of the total variance of far transfer. thinking styles which suit this teaching and learning
Again, we used Scheffe’s method and arrived at Table 6. method best are Judicial and the Liberal Styles in a
Table 6 shows that only the second and fourth traditional environment; but in the NetPBL environ-
treatments resulted in F 5 10.65, which is greater than ment, the best thinking style for near transfer is the
the critical value F 0 5 8.43. This means that only the Executive Style.
Legislative Group and the Mixed Group reached a Although the results of this study are somewhat
level of significance in far transfer. different from those found in the traditional environ-
ment, they still agree with basic concepts in the theory
of mental self-government. This theory suggests that
Conclusion and discussion
there is no criterion to judge whether one thinking
The conclusions of this study are as follows: style is better or worse than another, and that different

In the NetPBL environment, there are significant times and locations produce different judgements and
differences in the near transfer between the Executive evaluations. Because of the great discrepancies
Group and the Legislative Group. between the NetPBL and traditional approaches, the

In the NetPBL environment, there is no significant basic knowledge skills and personal interaction that
difference in far transfer among groups of different learners have to possess will be different. These
thinking styles. environmental differences might be the causes that

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39
38 C.-I. Lee & F.-Y. Tsai

lead to different conclusions. Moreover, different Finally, this study places great emphasis on the
cultures, gender, and ages are also important factors transfer of learning, which might be one of the causes
that influence thinking styles (Sternberg 1997, 2002; that somewhat differentiate the results of this study
Zhang 1999, 2002a; Zhang & Sternberg, 2002). Stu- from the other ones. Previous studies of thinking styles
dents in America, Europe, and Asia possess different mostly emphasized the relationship between thinking
thinking styles. A teaching and learning method that styles and academic achievement, cultures, and gender
most suits one particular thinking style might achieve differences (Sternberg 1997; Chen 2001; Kaufman
different results in various cultures. For example, in 2001). Among them, the number of researches on the
schools in Taiwan, students of the Executive Style relationship between thinking styles and academic
might get more encouragement than those of the achievement was more than abundant. But academic
Legislative Style; the same research conducted in achievement is not the same as the transfer of learning,
Europe and America may achieve different results. which is the independent variable emphasized in this
Besides, Sternberg (1997) takes the view that the study. Learning transfer has to be based on good
best way of group learning lies in each single group academic achievement (Hunter 1971). It is an ability
mixing students with different thinking styles. As the that is very important but not so easy to cultivate. In
results of this study show, the Mixed Group performs the NetPBL environment, relevant literature about the
better than the Legislative Group and the Judicial relationship between the transfer of learning and
Group in near transfer, while at the same time the thinking styles has not yet been undertaken. This
Mixed Group performs better than the Legislative might well be a topic worthy of further investigation.
Group in far transfer. Thus, before we group students
as a preparation of conducting cooperative learning,
we should take into consideration students’ thinking Limitations and suggestions
styles, just as in the case of a government needing the
This research combines constructivism and Stern-
mechanisms of execution, legislation, and judicature
berg’s theory of mental self-government with our
to produce the best effects. The results of this study
findings of the research, putting forward such fol-
not only verify the best grouping methods of the the-
lowing suggestions aimed at teaching. (1) It is more
ory of mental self-government, but also agree with the
proper for teachers to take thinking styles of students
viewpoints of other researchers that complementary
into consideration in teaching or teaching design for
roles in cooperative learning achieve better effects
NetPBL. For example, teachers should be aware of the
(Johnson & Johnson 1999).
thinking styles of students and make efforts to design
In addition, we reconfirm the thesis by the teachers
multiform teaching activities in order that students of
observing the students:
different thinking styles can benefit. (2) Teachers
should guide students to realize their own thinking
When students conduct PBL, they must set up sub- styles so that they master their characteristics and
questions, plan, and design experiment on their own
first. It is the characteristic of Legislative Style that
apply them to their own study or to cooperative
helps learners make a plan and design experiment sui- learning with other members. (3) The selected mem-
table for their team’s characteristic in a short time. The bers had better combine three different thinking styles
following task of execution (investigations and arran- after teachers’ grouping or by students’ seeking part-
ging and sharing artefacts), and the task of assessing
ners so as to induce students of different thinking
and analysing (analysing materials, monitoring experi-
ment) are the characteristics of Executive Style and styles to develop their merits and to enhance the
Judicial Style respectively. Therefore, a Mixed Group efficiency of individual or group study.
combining these three thinking styles can make better
use of the merits of members relatively and demonstrate
the results of cooperation. On the contrary, the group
with sole thinking style performs less efficiently. As far Acknowledgements
as the group of Judicial Style is concerned, it is superior
This research was supported in part by the National
in deciding on subquestions, planning, and designing
experiment. However, it is inferior to the Executive and Science Council of Republic of China under Grant No.
Judicial Groups in executing and analysing. NSC-90-2511-S-024-007.

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39
Internet project-based learning environment 39

References learning. Eductional Technology Research and Devel-

opment 46, 73–86.
Balakrishnan M. (2000) An investigation of the use of con-
Lin Y.C. (1999) The thinking styles of Hualien elementary
structivism and technology in project-based learning.
school students and the variables in stylistic develop-
Unpublished doctoral dissertation, University of Oregon.
ment. Unpublished Master Dissertation, National Hualien
Baldwin T.T. & Ford J.K. (1988) Transfer of training: a
Teachers College, Taiwan.
review and directions for future research. Personnel
Macaulay C. (2000) Transfer of learning. In Transfer of
Psychology 41, 63–105. Learning in Professional and Vocational Education (eds
Barry P.B., Fishman B.J., Krajcik J. & Marx R.W. (2001) V.E. Cree & C. Macaulay), pp. 1–26. Routledge,
Creating usable innovations in systemic reform: scaling London.
up technology-embedded project-based science in urban Marchaim U. (2001) High-school student research at Migal
schools. Educational Psychologist 35, 149–164. science institute in Israel. Journal of Biological Educa-
Broad M. (1997) Transferring Learning to the Workplace. tion 35, 178–182.
American Society for Training and Development, Alex- Marx R.W., Blumenfeld P.C., Krajcik J.S. & Soloway E.
andria, VA. (1997) Enacting project-based science. Elementary
Brown J.H. & Shavelson R.J. (1996) Assessing Hands-On School Journal 97, 341–358.
Science: A Teacher’s Guide to Performance Assessment. Murphy K.L. & Yakut G.D. (2001) Role plays, panel dis-
Corwin Press, California. cussions, and case studies: project-based learning in a
Chen C.H. (2001) Preferred learning styles and pre- web-based course. Paper presented at the Annual Meet-
dominant thinking styles of Taiwanese students in ac- ing of the American Educational Research Association,
counting classes. Unpublished doctoral dissertation. Seattle, WA, April 10–14 2001 (ERIC Document Re-
University of South Dakota. production Service No. ED 454 809).
Davis J.B. (1999) Improving Transfer of Learning in a Perkins D.N. & Salomon G. (1988) Teaching for transfer.
Computer Based Classroom. (ERIC Document Re- Educational Leadership 46, 22–35.
production Service No. ED 437 905). Sternberg R.J. (1988) Mental self-government: a theory of
Francisco C. & Elaine H.H. (2000) Learning and thinking intellectual styles and their development. Human De-
styles: an analysis of their interrelationship and influence velopment 31, 197–224.
on academic achievement. Educational Psychology 20, Sternberg R.J. (1994) Allowing for thinking styles. Educa-
413–430. tional Leadership 52, 36–40.
Grigorenko E.L. & Sternberg R.J. (1997) Styles of thinking, Sternberg R.J. (1997) Thinking Styles. Cambridge University
abilities, and academic performance. Exceptional Chil- Press, New York.
dren 63, 295–313. Sternberg R.J. (2002) Thinking styles and teachers’ char-
Han J. & Kamber M. (2001) Data Mining: Concepts and acteristics. International Journal of Psychology 37,
Techniques. Academic Press, San Diego, CA. 3–13.
Huang G.Q., Shen B. & Mak K.L. (2002) Participatory and Zhang L.F. (1999) Further cross-cultural validation of the
collaborative learning with TELD courseware engine. theory of mental self-government. Journal of Psychology
Journal of Professional Issues in Engineering Education 133, 165–181.
and Practice 128, 36–43. Zhang L.F. (2002a) Thinking styles and academic achieve-
Hunter M. (1971) Teach for Transfer. California Tip Pub- ment among Filipino students. Journal of Genetic Psy-
lications, El Segundo. chology 163, 149–164.
Johnson D.W. & Johnson R.T. (1999) Making cooperative Zhang L.F. (2002b) Thinking Styles: their relationships with
learning work. Theory Into Practice 38, 67–74. modes of thinking and academic performance. Educa-
Kaufman J.C. (2001) Thinking styles in creative writers and tional Psychology 22, 331–348.
journalists. Unpublished Doctoral Dissertation, Yale Zhang L.F. & Sternberg R.J. (1998) Thinking styles, abil-
University. ities, and academic achievement among Hong Kong
Krajcik J.S., Czerniak C.M. & Berger C. (1999) Teaching university students. Educational Research Journal 13,
Children Science: A Project-based Approach. McGraw- 41–62.
Hill, Boston, MA. Zhang L.F. & Sternberg R.J. (2002) Thinking styles and
Laffey J., Tupper T., Musser D. & Wedman J. (1998) A teachers’ characteristics. International Journal of Psy-
computer-mediated support system for project-based chology 37, 3–12.

& Blackwell Publishing Ltd 2004 Journal of Computer Assisted Learning 20, pp31–39