You are on page 1of 13

Educational Research Journal Vol. 14, No.

2, Winter 1999
Hong Kong Educational Research Association 1999
Cooperative CBI: The Effects of
Heterogeneous versus Homogeneous
Grouping, Student Ability and
Learning Accountability on
Achievement
Hong Kian-sam
Universiti Malaysia Sarawak
This study investigated the effects of cooperative group composition, stu-
dent ability, and learning accountability on achievement during computer-
based instruction. A total of 94 students aged between 14 to 15 years were
randomly assigned to heterogeneous and homogeneous dyads. Groups were
also assigned as having group or individual accountability for mastery of
lesson content. Cooperative dyads completed lessons on simple transfor-
geometry using a computer microworld. Students completed a post-
test five days later. There was an interaction effect between group composi-
tion and students' ability. Low ability students peiformed better in hetero-
geneous than in homogeneous groups. High ability students performed
slightly better in homogeneous than in heterogeneous groups. No signifi-
cant difference was found between individual and group accountability
groups.
Key words: cooperative learning; computer-based instruction;
mathematics achievement
Correspondence concerning this article should be addressed to Hong Kian-sam,
School of Education, University of Otago, P. 0. Box 56, Duredin, New Zealand.
302 Hong Kian-sam
Introduction
One frequently cited benefit of computer-based instruction (CBI) is the po-
tential to individualise instruction according to the needs of the learning
task, processing requirements and the current performance of the learner.
However, the costs of providing uniquely adaptive lessons delivered via
individual computers for each student are often prohibitive. Furthermore
the learning environment of such a learning scenario is inherently limited to
those strategies, explanations, and resources directly under computer control.
Students are not able to receive the varied explanations of their peers nor
gain the cognitive benefits associated with teaching among peers (Bargh &
Schul, 1980). Students who work individually for extended periods also
tend to become lonely, bored, or frustrated, resulting in lower achievement
motivation and a sterile and impersonal learning environment.
Educators interested in implementing CBI in education are concerned
with identifying models that maximise learning. One model that has gained
much attention involves the use of cooperative learning (Carrier & Sales,
1987; Johnson & Johnson, 1986; Johnson, Johnson & Stanne, 1985;
Meevarech, Stem & Levita, 1987; Webb, Ender & Lewis, 1986). To many,
cooperative learning has both strong intuitive appeal and compelling practi-
cal significance. The limited availability of computers in the classroom of-
ten mandates the use of a group model (Hannafin, Dalton & Hooper, 1987).
In most cases, individual learning at the computer may be both unnec-
essary and unwise (Hooper & Hannafin, 1988). That cooperative learning
methods can overcome many of the potential pitfalls of isolation while im-
proving students' achievement has been validated for CBI (Dalton, Hannafin,
& Hooper, 1989; Johnson, Johnson, & Stanne, 1985, 1986). Johnson and
Johnson (1989), in reviewing 180 studies comparing the achievement re-
sulting from cooperative versus individual learning, found an effect size of
0.66 in favour of cooperative methods.
Cooperative learning involves the selection of a number of students
(between two and five) to work in groups. The question of interest to educa-
Cooperative CBI
303
tors then is how should these cooperative groups be formed? How should
learners of varying abilities be grouped to maximise the benefits of coop-
erative grouping? A description of the learning phases, proposed by
Rummelhart and Norman (1978), may help to predict an effective model of
cooperative learning. They characterised learning as a process during which
the learner passes through three stages of understanding. In the first stage,
"accretion", the learner is able to discriminate between examples and non-
examples, but is unable to apply knowledge to new situations or provide in-
depth explanations. During the second stage, "restructuring", the learner
can transfer some learning but is still unable to provide deep understanding.
At the final stage, the learner enters the highest level of learning, "tuning",
and is able to solve novel problems, to work effectively under stress and to
provide deep explanation.
Based on this model, cooperative learning advocates typically recom-
mend that students are grouped heterogeneously, that is, group composition
is manipulated to include students with diverse experiences. Heterogeneous
grouping is encouraged for both affective and cognitive considerations. Stu-
dents encounter wider diversity in heterogeneous than in homogeneous
groups. Thus, heterogeneous grouping is more likely to improve inter-per-
sonal attraction among group members and help dismantle social barriers
(Johnson & Johnson, 1989). Heterogeneous ability grouping benefits both
high and low ability students. Less able or disadvantaged students receive
more instructional support and regulation from their partners than from the
classroom teacher, are more actively involved, and may observe their part-
ners' learning strategies (Swing & Peterson, 1982). Furthermore, low abil-
ity students in homogeneous groupings are likely to flounder in an environ-
ment that requires group members to explain cognitive information.
Concurrently, more able students may also benefit cognitively from explain-
ing lesson concepts to their partners and from the opportunity to practise
important skills in both heterogeneous or homogeneous groups (Bargh &
Schul, 1980; Mayer, 1984).
304 Hong Kian-sam
Despite the potential social benefits, the cognitive effects of heteroge-
neous ability grouping have not been established. Some research indicates
that heterogeneous grouping of high and low ability students supports the
needs of one group at the expense of another. Beane and Lemke (1971)
found that heterogeneous ability grouping improved the achievement of the
most able group members at the expense of the least able. In contrast, Hooper
and Hannafin (1988) indicated that heterogeneous grouping increased the
achievement of low ability students by approximately 50% compared to
their homogeneously grouped peers. However, homogeneous grouping in-
creased the achievement of high ability students by approximately 12%
compared to their heterogeneously grouped counterparts. This suggests sig-
nificant payoffs for low ability students from heterogeneous groupings but
potential decrements in the performance of their high ability cooperative
learning partners.
The effectiveness of cooperative learning is often attributed to interac-
tion among group members (Webb, 1989), but little is known about the
relationship between intra-group interaction and achievement. Webb, Ender
and Lewis (1986) indicated that the nature of intra-group cooperation is
potentially of greater importance than group composition per se. It is im-
portant to determine how group composition influences intra-group inter-
action and to develop methods that promote successful interaction.
One method to promote interaction involves increasing individual
accountability, wherein each group member must demonstrate mastery of
content embedded in the instruction. Compared with deriving a "team
response", where less able students might simply defer to those who are
more able, or more able students may attempt to dominate, individual ac-
countability may promote superior interaction, qualitatively and
quantitatively. This method may be used to isolate and overcome potential
learning problems within a group and to provide an additional incentive to
cooperate (Hooper, Ward, Hannafin, & Clark, 1989). This method may re-
Cooperative CBI
305
duce the damaging "free rider" and "sucker" effects (Kerr, 1983; Kerr &
Bruun, 1983) by motivating more able group members to provide help, and
less able members to invest sufficient mental effort to master instruction.
Malaysian schools have an estimated enrolment of 5.8 million students
(Ministry of Education, 1997). Aiming for individualized CBI will be unat-
tainable and unrealistic, and as Hooper and Hannafin (1988) state, unneces-
sary and unwise. Cooperative CBI is an approach worth exploring to enable
Malaysian students to benefit from the use of computers in schools. In the
Malaysian educational setting, two papers that support the use of coopera-
tive learning and cooperative CBI were authored by Gan (1992, 1994). Gan
argued that the use of cooperative learning is suitable in the context of the
integrated curriculum practised in the Malaysian educational system and
the multiracial nature of Malaysian classrooms. Gan also successfully de-
veloped and used environmental education courseware for cooperative learn-
ing activities in Malaysian classrooms.
Purposes of Study
The purposes of this study were to examine (a) the effects of heterogeneous
and homogeneous group ability composition on achievement; and (b) the
effects of strategies requiring different levels of performance accountability.
The achievement and interaction of high and low ability students were com-
pared in heterogeneous and homogeneous groups featuring either individual
or group performance accountability. It was predicted that low ability stu-
dents would demonstrate higher achievement in heterogeneous than in ho-
mogeneous groups. High ability students in heterogeneous groups were pre-
dicted to demonstrate equal or better achievement as compared to those in
homogeneous group. It was also predicted that achievement would be higher
for students in individual accountability groups than for those in group ac-
countability groups.
306 Hong Kian-sam
Method
A sample of 94 Form Two students (aged 14 to 15 years) from an urban
school in Malaysia participated in the study. Only students with uniformly
high or low performances on both the Form One final examination
(mathematics test) and study pre-test were included. High ability students'
were defined as students having scores at above the 70th percentile for the
mathematics test in the Form One final examination and above the mean on
a pre-test specifically designed for the study; low ability students scored at
or below the 40th percentile and below the mean on the pre-test.
Materials
Measurement Instruments
The pre-test and post-test are the same measurement instrument but with
different sequencing of the test items. The measurement instrument com-
prised 30 multiple choice items on simple translation, reflection and rota-
tion in transformational geometry. The use of multiple choice items for trans-
formational geometry for students aged 14 to 15 years was supported by
Perham's (1978) research. The measurement instrument was modelled by
constructing a test specification table based on the Malaysian Lower Sec-
ondary Mathematics Syllabus. Face validity for the measurement instru-
ment was done by a mathematics education lecturer from SEAMEO
RECSAM (Regional Centre for Science, Mathematics and Technology
Education) and two practising mathematics teachers in the participating
school. The Cronbach Alpha reliability for the measurement instrument was
0.93.
Cooperation Training
Training was designed to facilitate effective intra-group interaction and
cooperation. The first training session emphasised peer awareness. Students
were introduced to the game "Broken Circles" (Cohen, 1986). Broken Circles
is a puzzle that cannot be completely solved unless students sacrifice indi-
Cooperative CBI
307
vidual success for the good of the group. During this training session, groups
of four to six members were formed. Each member was given an envelope
that contained pieces of a circle. None were given pieces that were initially
adequate to construct a complete circle, and the objective of this activity
was for every group member to create a circle. Interaction was limited. Stu-
dents were allowed to offer pieces, but talking was not permitted and taking
game pieces from other group members was forbidden.
The aim of the second training session was to promote oral
summarisation between group members. Students in dyads completed three
tasks involving identifying and summarising rules to a partner. The first
task required students to work together to learn a large number which repre-
sented the square of a series of natural numbers (e.g. 149,162,536 repre-
sents "1 ", "4", "9", "16", "25", "36" - the squares of the number 1 through
6). The second task required that students identify the number of rectangles
or squares embedded within a complex figure. Students were told to iden-
tify the rules that governed each task and then take turns summarising the
rules to each other. When a task was completed, students were asked to
explain the rules. Between each task, feedback concerning the appropriate-
ness and effectiveness of student behaviour was provided.
CBI Lesson Content
A computer microworld TRANSFORM which simulates the transforma-
tion geometry lesson was developed by the researcher. A microworld is a
computer-based learning environment that embodies mathematical concepts
in a context which is engaging to the learner and which allows a certain
degree of self-directed exploration or discovery of the implicit ideas and
processes (Edwards, 1985). TRANSFORM enables students to draw ob-
jects and perform simple transformations (i.e., translation, reflection and
rotation). Students view the results of their actions on the computer screen.
TRANSFORM was designed to fit the requirements of the Malaysian school
syllabus on transformational geometry for 14-to 15-year-old students in terms
of appearance, terminology and symbols. Working in dyads, students ex-
308 Hong Kian-sam
plored this microworld guided by simple cooperative-based worksheets
(Hong, 1996).
Design and Data Analysis
The study employed a 2 x 2 x 2 factorial design. The between subjects
factors included Ability (high, low); Cooperative Group Composition
(heterogeneous, homogeneous); and Accountability (group, individual). The
dependent measure, post-test scores were analysed through ANOVA pro-
cedures using a significant level of 0.05.
Procedures
The pre-test was administered to all potential students in the target school
to identify those with high or low mathematics ability. Students thus identi-
fied then completed three 30-minute training sessions during a two-day
period. Students were informed they would work in dyads. Using stratified
random sampling, heterogeneous and homogeneous ability groups were
established. Heterogeneous groups contained one high ability student and
one low ability student. Homogeneous ability groups contained two high
ability students or two low ability students. Each group was assigned to a
computer and completed either the group or individual accountability ver-
sion of the CBI lesson. Students from the individual accountability dyads
were each given quiz booklets to be completed without help from their
prutners; students in group accountability condition were given a single quiz
booklet to complete cooperatively. Students were allocated eight 40-minute
sessions to follow the instructions and practice segments of the lesson. Sub-
jects received the post-test five days after the completion of the CBI lessons.
Results
Post-test means and standard deviations for each treatment are shown in
Table 1. The results of the corresponding AN 0 VA are found in Table 2. The
overall post-test means of the high ability (24.604) and low ability (12.163)
Cooperative CBI
309
groups were significantly different, E(1,86) =811.942, 11 < 0.0005. The overall
post-test means of the heterogeneous (19.542) and homogeneous (17.225)
groups were significantly different, E(1,86) = 28.151,11 < 0.0005. No sig-
nificant differences were found for responsibility, .E(l,86) = 0.499, 11 = 0.
482.
Table 1 Post-test means and standard deviations
Ability Accountability Homogeneous Heterogeneous Total
High IRM 25.250 24.000 24.625
SD 2.4921 2.000 2.246
(n) 12 12 24
GR M 24.750 24.417 24.583
SD 1.658 1.677 1.668
(n) 12 12 24
Total M 25.000 24.208 24.604
SD 2.075 1.839 1.957
(n) 24 24 48
Low IRM 9.000 14.667 11.833
SD 1.954 1.875 1.915
(n) 12 12 24
GRM 9.900 15.083 12.492
SD 3.348 1.621 2.485
(n) 10 12 22
Total M 9.450 14.875 12.163
SD 2.651 1.748 2.199
(n) 22 24 46
Overall Total M 17.225 19.542 18.383
SD 2.363 1.793 2.078
46 48 94
IR: Individual Responsibility GR: Group Responsibility
A significant interaction effect was detected between student ability
and type of grouping E(1,86) = 50.678, 11<0.00005. Follow-up contrast re-
vealed that the mean post-test scores of low ability students in heteroge-
neous groupings (mean= 14.875) was significantly higher than those in
homogeneous groupings (mean= 9.450); nonetheless the mean post-test
310 Hong Kian-sam
scores for high ability students in heterogeneous groupings (mean= 24.
208) was not significantly different from those in homogeneous groupings
(mean= 25.000).
No other significant differences were found for other levels of interaction,
i.e., ability x responsibility, E(l ,86) = 0.643, n. = 0.425; group x
responsibility, E(l,86) = 0.062, n. = 0.805; and ability x group x
responsibility, E(1,86) = 0.643, IL= 0.425.
Discussion
The predicted interaction between ability and group composition was
significant. Ability grouping appears to influence achievement. Learning
was most efficient for homogeneously grouped high ability students and
least efficient for homogeneously grouped low ability students; apparently
low ability students benefited from, but slightly slowed, the progress of
high ability students. These findings tend to support previous cooperative
learning studies that cooperative learning poses little risk to more able tu-
tors (Hopper & Hannafin, 1988).
In this study, students who collaborated on the quizzes did not score
higher on the post-test than those who completed the quizzes individually.
This finding suggests that the "free rider" effects, which often occur in group
learning, were absent in this study.
Recommendations for Future Research
Four recommendations for future research should be noted. First, group
learning was not compared to individualised instruction in this study. Fu-
ture research should include control groups to compare the effects of ability
grouping across groups and individual treatments. Second, findings from
this group may not be generalised to larger groups. Free-riding is more apt
to occur when groups include more than two persons; thus the issue of so-
cial dilemmas may be more important in larger groups. Researchers should
also examine whether free-riding will be more prevalent during longer treat-
Cooperative CBI
311
ments as free-rider effects may be less prevalent during a short experimen-
tal study due to novelty effects. Thirdly, in this study heterogeneous groups
consist of high and low ability students. Controversy and argumentation
increase as intra-group heterogeneity increases (Johnson & Johnson, 1989).
The results may not be generalisable to other kinds of heterogeneous groups,
i.e., groups containing students of high and average abilities, average and
low abilities, or high and low abilities. Further research is needed to exam-
ine the effects of these group compositions. Finally, what are the precise
cooperative processes during the cooperative CBI: What do students really
do in groups? Do their behaviours differ depending on their partner's ability?
These questions can only be answered if further research involving video
data is carried out.
References
Bargh, J. A., & Schul, Y. (1980). On the cognitive benefits of teaching. Journal of
Educational Psychology, 72, 593-604.
Beane, W. E., & Lemke, E. A. (1971). Group variables influencing the transfer of
conceptual behavior. Journal of Educational Psychology, 62, 215-218.
Carrier, C. A., & Sales, G. C. (1987). Pair versus individual work on the acquisition
of concepts in a computer-based instructional lesson. Journal of Computer-
Based Instruction, 14, 11-17.
Cohen, E. G. (1986). Designing groupwork: Strategies for the heterogeneous
classroom. New York: Teachers College Press.
Dalton, D. W., Hannafin, D. W., & Hooper, S. (1989). The effects of cooperative
learning strategies on achievement and attitudes during interactive video. Jour-
nal of Computer-Based Instruction, 17, 8-16.
Edwards, L. D. (1985). The design and analysis of a mathematical microworld.
Journal of Educational Computing Research, 12(1), 77-94.
Gan, S. L. (1992). Integrated curriculum and cooperative learning in Malaysian
multiracial classrooms. Paper presented at the World Council for Curriculum
and Instruction Seventh Triennial World Conference on Education, Cairo.
Gan, S. L. (1994). Developing and using courseware for cooperative learning ac-
tivities in the classroom. Paper presented at Asia Pacific Information Technol-
ogy in Training and Education 1994 Conference, Brisbane, Australia.
312 Hong Kian-sam
Hannafin, M. J., Dalton, D. W., & Hooper, S. R.(1987). Computers in education:
10 myths and 10 needs. Educational Technology, 27(10), 8-14.
Hong, K. S. (1996). The effects of group formation on achievement using Coopera-
tive CBI. Unpublished Master's Thesis, Universiti Sains Malaysia, Penang,
Malaysia.
Hooper, S., & Hannafin, M. J. (1988). Cooperative CBI: The effects of heteroge-
neous versus homogeneous grouping on the learning of progressively complex
concepts. Journal of Educational Computing Research, 4(4), 413-424.
Hooper, S., Ward, T. M., Hannafin, M. J., & Clark, H. T. (1989). Factors influenc-
ing small group learning in a college age population. Journal of Computer-
Based Instruction, 16, 102-109.
Johnson, D. W., & Johnson, R. T. (1986). Computer-assisted cooperative learning.
Educational Technology, 26(1), 12-18.
Johnson, D. W., & Johnson, R. T. (1989). Cooperation and competition: Theory
and research. Edina, MN : Interaction Book Company.
Johnson, D. W., Johnson, R. T., & Stanne, M. B. (1985). Effects of cooperative,
competitive, and individualistic goal structures on computer-assisted instruction.
Journal of Educational Psychology, 77, 668-677.
Johnson, D. W., Johnson, R. T., & Stanne, M. B. (1986). Comparison of computer-
assisted cooperative, competitive, and individualistic learning. American Edu-
cational Research Journal, 23, 382-392.
Kerr, N. L. (1983). Motivation losses in groups :A social dilemma analysis. Jour-
nal of Personality and Social Psychology, 45, 819-828.
Kerr, N. L., & Bruun, S.E. (1983). The dispensability of mental effort and group
motivation losses : Free rider effects. Journal of Personality and Social
Psychology, 44, 78-94.
Latane, B., Williams, K., & Harkins, S. (1979). Many hands make light work: The
causes and consequences of social loafing. Journal of Personality and Social
Psychology, 37, 822-832.
Mayer, R. E. (1984). Aids to text comprehension. Educational Psychologist, 19(1),
30-42.
Mevarech, Z. R., Stem, D., & Levita, I. (1987). To cooperate or not to cooperate in
CAl : That is the question. Journal of Educational Research, 80, 164-167.
Ministry of Education ( 1997). Smart school flagship application: The Malaysian
smart school- A conceptual blueprint. Malaysia: Government of Malaysia.
Cooperative CBI
313
Perham, F. (1978). An investigation into the effect of instruction for the acquisition
of transformational geometry concepts in first grade children and subsequent
transfer to general spatial ability. In R. Lesh (Ed.), Recent research concerning
the development of spatial and geometric concepts. The Ohio State University:
ERIC Clearinghouse for Science, Mathematics and Environmental Education.
Rummelhart, D., & Norman, D. (1978). Accretion, tuning and restructuring. In J.
Cotton & L. Klatzky (Eds.), Semantic factors in cognition (pp. 37-53). Hillsdale,
NJ: Lawrence Erlbaum.
Swing, S. R., & Peterson, P. L. (1982). The relationship of student ability and small-
group interaction to student achievement. American Educational Research
Journal, 19, 259-274.
Webb, N. M. (1989). Peer interaction and learning in small groups. International
Journal of Educational Research, 13(10), 21-39.
Webb, N. M., Ender, P., & Lewis, S. (1986). Problem solving strategies and group
processes in small groups learning computer programming. American Educa-
tional Research Journal, 23, 243-261.

You might also like