Journal Educational

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

The Effects of Cooperative Learning on Junior

High School Students’ Behaviours, Discourse and


Learning During a Science-Based Learning
Activity

ROBYN M. GILLIES
School of Education, The University of Queensland,
Brisbane, Australia

ABSTRACT The study investigated the effects of structured and


unstructured cooperating groups on students’ behaviours, discourse
and learning in junior high school. One hundred and sixty-four grade 9
students participated in the study. The students were videotaped as
they worked in three to four person, mixed-gender and ability groups
on a science-based categorization activity. The results show that the
students in structured cooperating groups demonstrated more co-
operative and helping behaviours such as giving more elaborated help
and guided directions to assist understanding than their peers in the
unstructured groups. Moreover, they demonstrated more complex
thinking and problem-solving skills both in their discourse and their
responses on the follow-up learning probe. These findings are dis-
cussed in the context of the importance of structuring cooperative
learning experiences if students are to attain the benefits widely
attributed to this approach to learning.

KEY WORDS: cooperative learning; cooperative pedagogy; small group


work; teacher discourse; teacher perceptions

There have been a plethora of studies over the last 30 years that have
documented the benefits of cooperative learning for students from
kindergarten through to college level. These benefits have covered the
academic, social and emotional domains and include the development
of more positive and supportive relationships and greater psychological
health and well-being for students who participate in cooperative learn-
ing than those who experience the more traditional, competitive and
individual approaches to instruction (Johnson and Johnson, 2002;
Shachar and Sharan, 1994; Stevens, 2003). Cooperative learning has

School Psychology International Copyright © 2008 SAGE Publications (Los


Angeles, London, New Delhi and Singapore), Vol. 29(3): 328–347.
DOI: 10.1177/0143034308093673

328

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

also been used to promote effective social communication skills among


delinquent adolescent girls who were ineffective communicators
(Rutherford et al., 1998), to resolve conflicts in early, middle and high
school settings (Johnson et al., 1997; Stevahn et al., 2000) and as a
strategy for promoting multicultural inclusion (Johnson and Johnson,
2000; Slavin and Cooper, 1999). In addition, cooperative learning
experiences have assisted in the acceptance of children with disabilities
by their non-disabled peers and have been shown to enhance small
group interactions among children with developmental disabilities
(Kamps et al., 1994) and learning difficulties (Gillies and Ashman,
2000; McMaster and Fuchs, 2002). In fact, it can be argued that it is no
longer necessary to defend cooperative learning as a teaching strategy
that promotes learning. The evidence of its effectiveness is unequivocal.

Cooperative learning: what is it?


Cooperative learning involves students working together to accomplish
shared learning goals (Johnson and Johnson, 2003). When children
cooperate, they learn to listen to what others have to say, give and
receive help, share ideas, clarify differences and construct new under-
standings (Johnson and Johnson, 2003; Webb et al., 1995). It is this
sense of working together that promotes group cohesion and helps the
group to develop its identity as a unit with common goals to attain
(Slavin, 1996). From a socio-cultural perspective, the group context
enables students to interact with others and think about issues in ways
that they had never considered previously. In so doing, the information
and ideas exchanged are transformed and appropriated so they become
part of their new ways of knowing and doing or generating knowledge
(Palincsar, 1998). By engaging in reciprocal interactions with others,
children learn to use language differently to explain new experiences
and new realities and, in so doing, they learn to find new functions for
language in thinking and feeling (Mercer et al., 1999). The interactions
that occur are multi-directional as students respond to explicit and
implicit requests for help and scaffold their responses to facilitate
understanding and learning in others (Gillies and Ashman, 1998).
Furthermore, children are often more aware than their teachers of
what other students do not understand and will direct their attention
to the important features of a problem that needs to be solved, and, in
so doing, often use language that can be easily understood by the recip-
ient (Webb and Farivar, 1999). When children do this, they engage in
teaching within the learner’s zone of proximal development (Palincsar,
1998; Vygotsky, 1978). Others have observed similar teacher–learner
interactions when students work cooperatively together (King, 1999)
and Gillies and Ashman (1998) reported that such interactions were

329

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

observed among first grade children as they worked in cooperating


groups. Children can and do actively facilitate each other’s learning
and understanding as they engage in reciprocal interactions around
common, cooperative tasks (Cowie and Berdondini, 2001; Palinscar
and Herrenkohl, 2002).

Task-related interactions
The frequency of task-related interactions in groups is important for
productivity and follow-up gains on content referenced tests and con-
ceptual development (Cohen, 1994; Cohen et al., 1997). Shachar and
Sharan (1994) found that children in cooperative groups expressed
themselves more frequently, used more words per turn of speech and
engaged in more intellectually valuable conversations than their peers
who worked in groups where they had not been taught to cooperate.
Moreover, when children work in cooperative groups they engage in
more task-related interactions than their peers who work either in
whole class settings or in informal cooperative groups, and it is the
frequency of these interactions that have significantly predicted higher
learning outcomes for students (Cohen et al., 2002; Gillies, 2003;
Gillies and Ashman, 1998).
King (2002) found that higher-level thinking can be promoted during
group discussions when students are taught to ask specific thought-
provoking questions designed to encourage their peers to think more
carefully about the topic under discussion and provide more elabora-
tive responses, explanations or justifications for their respective views.
When students learn to ask and answer these particular questions,
they learn to monitor and regulate their understanding of the material
and to extend their learning beyond the topic under discussion to
construct new knowledge. Similarly, Abram et al. (2002) found that
when students were provided with specific criteria that required them
to critically evaluate their own work, they engaged in more evaluative
talk than their peers who were not provided with such criteria.
Furthermore, there were significant correlations between the use of
evaluative talk and individual scores on the content and conceptual
understanding of follow-up essays. These essays were designed to
determine how the children were making connections between what
they had learned from the written components (i.e. supporting materi-
als) of their work unit and their small group discussions. In effect, it
appears that the benefits of group discussion can be enhanced if
students are taught to engage in critical and constructive discourse
with each other where they learn to ask challenging questions so their
peers are encouraged to justify and evaluate their understanding of the
problem at hand.

330

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

Implementation difficulties
Despite the benefits that accrue to students from cooperative learning,
the use of this nontraditional approach to instruction appears to be
limited, with many schools still tending to prescribe how students will
learn, the topics to be covered in each subject, the number of periods
per subject per week and the types of tests to be administered to moni-
tor student progress (Sharan et al., 1999). In such schools, there is little
room for innovation in how teachers teach and how students learn.
Teachers are viewed as experts who instruct while students are seen as
passive recipients of knowledge (Sharan et al., 1999).
In an investigation of elementary students’ perceptions of different
classroom teaching methods used in science and mathematics, Race
and Powell (2000) found there was a decline in the use of non-
traditional approaches such as cooperative learning as students moved
into the higher grades and these perceptions tended to mirror students’
performance levels in mathematics (science performance levels are
not reported). Furthermore, teachers in the higher grades reported
less favourable attitudes towards using non-traditional teaching
approaches in their classrooms than teachers in the lower grades.
Baines et al. (2003), in a study of grouping practices in the UK,
reported that by secondary school students rarely worked in small
groups; most worked in dyads or of groups of 11 or more members with
little autonomy over the group size, task or how members were to inter-
act. Grouping practices were aimed at maintaining control, keeping
students on-task and maximizing teacher-directed learning.
In a study that examined the prevalence, conceptualization and form
of cooperative learning used by elementary teachers in the US, Antil et
al. (1998) found that few teachers were employing recognized forms of
cooperative learning in their classrooms even though all had indicated
that they employed cooperative learning daily in different subjects.
Similar observations have been made by Gillies (2003) about teachers’
practices in Australian schools.
A reluctance to embrace cooperative learning may be due, in part, to
the challenge it poses to teachers’ control of the classroom environ-
ment, including instructional and organization changes that may need
to be made to accommodate this approach to learning as well as the
personal commitment teachers need to make to sustain their efforts
(Kohn, 1992). It may also be due to a lack of understanding of how to
embed cooperative learning into the curriculum to promote cooperative
investigation, discussion and joint problem-solving (Johnson and
Johnson, 2003; Sharan et al., 1999). There is no doubt that many
teachers group students for convenience rather than ensuring that the
key elements that identify cooperative learning are present (Baines et
al., 2003; Gillies, 2003).

331

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

Structuring cooperative learning


Although there are numerous benefits that students derive from co-
operative learning, it is also clear that placing students in groups and
expecting them to work together will not always promote cooperation
and learning. It is only when groups are structured so students under-
stand how they are expected to cooperate that the potential for
cooperative learning is maximized (Johnson and Johnson, 1990). This
includes ensuring that the group task is established so that all mem-
bers realize that they are required to contribute and to assist others to
do likewise. It also includes ensuring that students are taught the
interpersonal and small-group skills that are required to help students
communicate effectively with their peers, manage conflict, allocate
resources fairly and make decisions democratically. When these ele-
ments have been embedded into the small group structure, students
are more likely to feel included and accepted as part of the ‘group’, and
this, in turn, provides the impetus for them to feel motivated to achieve
and contribute to both their own and the group’s goals (Johnson and
Johnson, 2003).
However, despite the importance of structuring cooperative learning,
many teachers still continue to establish small groups where these
elements are often not fully established, that is, many of the groups are
constructed in an ad-hoc or unstructured way where students are not
linked interdependently around the task, nor are they taught the social
skills needed to promote constructive discussion and learning. In short,
the groups are unstructured.
The present study involved junior high school students who worked
in either structured or unstructured small groups (as defined above)
during inquiry-based science. Inquiry-based science learning involves
students working cooperatively to generate their own hypotheses,
construct their own working theories, critically evaluate different
conceptions and generate and search for new knowledge and under-
standings (Veermans et al., 2005), based on the assumption that
children are more motivated to learn when they are encouraged to be
active participants in the learning process, investigate problems that
challenge their curiosity and think creatively as the work towards com-
monly agreed conclusions (Turner and Patrick, 2004). The purpose of
the study was to investigate if there are differences in how students
behave, interact and learn when they work in groups that are struc-
tured as opposed to those that are unstructured during a science-based
learning activity.

332

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

Methods

Design of the study


This is a comparative study of students’ behaviours, discourse and
learning in classes where teachers have established either structured
or unstructured cooperative groups during inquiry-based science
classes in junior high school. Identification of the structured and
unstructured groups is based on teacher and student reported use of
cooperative learning, discussions with key administrators (e.g. princi-
pal, deputy principal) and examination of school policy documents
(discussed below).

Participants
One hundred and sixty-four grade 9 students (77 students in the struc-
tured condition and 87 students in the unstructured condition) from six
high schools in Brisbane, Australia participated in the study. While
most groups consisted of four students, student absenteeism on the day
the research team visited to videotape meant that seven groups in the
structured condition (n = 21 groups) and five groups in the unstruc-
tured condition (n = 23 groups) consisted of only three students.
Five of the schools included in this study were identified as having a
similar socioeconomic and demographic profile (i.e. a similar mixture of
social, cultural and economic diversity among their students). In the
one school that had a different socioeconomic and demographic profile,
nearly one-third of the children came from single-parent families and
nearly 10 percent of the children were transient and changed school
regularly (this school had participated in extensive professional devel-
opment on cooperative learning). However, there were no significant
differences between the schools prior to the start of the intervention
in the structured and unstructured conditions (discussed below) in
teachers’ ratings of students’ achievements in science (ANOVA), F(1,
160) = 2.12, p > 0.05 (Structured condition M = 3.59, SD = 0.97;
Unstructured condition M = 3.81, SD = 0.89) on a five-point rating scale
(1 = unsatisfactory achievement to 5 = very high achievement) (a previ-
ous school x science achievement one-way ANOVA was not significant,
F(5, 161) = 1.77, p > 0.05 indicating there were no difference between
the schools). The teachers’ ratings were based on the children’s test
results in science during the first term in grade 9. All teachers (three
from each condition) were given very specific guidelines on how to rate
the children’s performance to ensure consistency in ratings across
schools (i.e. on a scale from 1–10, a rating of 9 or 10 designated very
high achievement and was indicative of children being able to success-
fully complete problem-solving tasks that represented an extension of
knowledge or information that they had been taught or would have had

333

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

the opportunity to learn from their involvement in science activities in


their classroom; a rating of 4–6 designated satisfactory achievement
and was indicative of children who were able to understand and apply
basic information to problem-solving tasks).
Stratified random assignment occurred within each class so that
each four-person group consisted of one high-achieving student (top
quartile), two medium-achieving students (quartiles 2 and 3) and one
low-achieving student (bottom quartile). This assignment corresponds
closely to decisions often used in previous research for classifying
students into ability levels (e.g. Gillies and Ashman, 1998; Webb et al.,
1997). Previous research has shown while there is no evidence that one
form of grouping is better than another, low ability children do benefit
from interacting with higher ability peers and high ability children are
not disadvantaged by working in mixed ability groups (Lou et al.,
1996).

Structured versus unstructured cooperative learning


All schools that participated in the study indicated that they encour-
aged their teachers to use a wide range of non-traditional practices,
including cooperative learning, to teach their students. However, dis-
cussions with the teachers and students and key school administrators
(i.e. principal, deputy principal) and examination of school policy docu-
ments revealed that cooperative learning was not a widespread
teaching strategy. For example, some schools in the study had partici-
pated in extensive professional development activities that included
training in how to establish cooperative learning groups in their class-
rooms (i.e. structuring cooperative groups so that the key elements of
positive interdependence, promotive interaction, individual account-
ability and group processing were present and students were trained in
the social skills needed to promote cooperation), understanding and
using different learning styles in teaching and working collaboratively
with colleagues. Lopata et al. (2003) found that cooperative learning is
more likely to be used by teachers when they have participated in staff
development designed to provide them with the background knowledge
and skills required to implement this approach to learning in their
classrooms.
Three schools had had a strong commitment to cooperative learning
because they encouraged their teachers to embed structured coopera-
tive learning (as discussed above) in their curricula and to provide
students with opportunities to participate in cooperating group activi-
ties at least once a week. It should be noted that the students in these
schools were also participating in cooperative learning activities in
their English and mathematics classes, so most students were involved
in cooperating groups at least once or twice a week, although they

334

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

worked in different groups with different students in these subject


classes. Three schools, in contrast, did not implement structured co-
operative learning (as discussed above) in their classes, nor did
students participate in cooperating groups on a regular basis.
Cooperative learning was defined as involving three or more stu-
dents who worked together on a common group activity that required
them to contribute to the task (positive task interdependence), help
each other (by using the appropriate interpersonal and small-group
skills, asking intelligent questions and providing detailed responses to
questions) and promote each other’s learning (Johnson and Johnson,
2000; King, 2002). In the schools that had a strong commitment to
cooperative learning, small groups were established so that these key
elements of cooperative learning were evident. In effect, the small
group learning activities were structured.
In contrast, in the schools that did not have a strong commitment to
cooperative learning, the students were involved in small group activi-
ties in their classes on an irregular basis and the key elements that
identify cooperative learning were not evident; that is, small groups in
these schools were unstructured.

Group problem-solving task


In the group problem-solving task, the students were required to apply
the principles of classification for all living things that they that they
had been studying in science to a non-science activity. The purpose of
the problem-solving task was to see if the students were able to gener-
alize and extrapolate these principles to a non-science activity where a
system of classification was needed to help make sense of specific infor-
mation. In this instance, the students were required to organize their
knowledge of local television programmes into specific categories and
sub-categories and to identify additional sub-categories on the basis of
specific attributes – all within a 10 minute time limit.
The students were given 42 plastic-covered name cards of different
television programmes which they used to structure the classification
tree. For example, the first activity required the students to sort the
programmes into national and non-national programs. Using this
information, they were then required to identify and sort the pro-
grammes into a ratings classification. Once they had completed this
task, they were then required to identify five additional categories that
sorted the programmes into classes of similar programmes and be
prepared to explain and justify their categorizations. This was followed
by an activity that required them to identify attributes of programmes
within each of these sub-categories so they, in turn, were distinguished
from others. The final activity required the students in the group
to draw the classification tree and label the different component

335

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

categories and sub-categories and be prepared to share their product


with the wider class group. This activity was specifically developed for
the group problem-solving video session because it was original and it
sought to determine if the students could extrapolate the principles of
scientific classification to a non-science activity.

Measures
Behavioural observations. The schedule used to gather information on
students’ behaviour states during recorded small group sessions
was originally developed by Sharan and Shachar (1988), but has been
modified by Gillies and Ashman (1996, 1998) to focus on four mutually
exclusive behaviour states: (a) cooperative behaviour (task-orientated
group behaviour, socially engaged in the group task, listening to
others); (b) noncooperative behaviour (competitive behaviours designed
to exclude others, expression opposition or criticism); (c) individual
task-oriented behaviour (working alone on the task) and (d) individual
non-task behaviour and confusion (not participating in group activities
and not working individually). Momentary time sampling was used to
record the occurrence of behaviour within each category at 10 second
intervals for group members over a period of 10 minutes. Only the
behaviour that was observed at each 10 second interval for the student
who was being observed was recorded. In total, 15 observations were
recorded for each group member.
Verbal interactions. This schedule, which is based on one developed
by Webb (1985) and modified by Gillies (2003), was designed to compile
information on student verbal interactions during recorded group
sessions. There were nine interaction variables (Verbal Interactions):
(a) directives (i.e. verbal instruction); (b) directives with physical
prompt (i.e. verbal instruction with hand gestures or body prompts)
(c) unsolicited explanation (i.e. providing elaborated and detailed help
when it is not requested); (d) unsolicited terminal (i.e. gives short
answer response that is not detailed or elaborated); (e) positive inter-
ruption (i.e. interjects to help in a constructive way); (f) negative
interruption (i.e. designed to interrupt others); (g) solicited explanation
(i.e. provides detailed and elaborated help in response to request for
assistance); (h) solicited terminal (responds with short response or help
on request) and (i) nonspecific interactions (all other verbal inter-
actions that do not fit into one of the above categories). Verbal
interactions were coded according to frequency across the recorded
group session. Two observers, blind to the purposes of the study coded
a common five hours of videotape. Interobserver reliability ranged from
92 to 100 percent across the Behaviour States and 80 to 93 percent
across the Verbal Interactions.

336

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

Cognitive language strategies


This schedule focused on the cognitive language strategies used by the
students during their group problem-solving activitiy and was origin-
ally developed by Sharan and Shachar (1988) but modified by Gillies
and Ashman (1998). The cognitive language strategies were more
content oriented that the verbal interactions schedule which focused
more on the frequency of the use of specific verbal interactions. The six
cognitive category variables were: (a) repeats information (i.e. repeats
information in a mechanical way, basic repeating of facts); (b) unstruc-
tured idea (i.e. talks out loud but ideas are loosely organized);
(c) concrete idea (i.e. states a fact related to the topic); (d) explanation
with evidence or justification (i.e. states reason); (e) generalizes infor-
mation (i.e. applies ideas to other situations or states a general
principle); (f) evaluative comment (i.e. considers multiple sources of
information to arrive at an answer). Once again, observers, blind to the
purposes of the study, coded a common three hours of videotape
and inter-observer reliability ranged from 83 to 89 percent across the
cognitive language strategies.

Science probe questionnaire


A probe consisting of six levels of questions based on Bloom’s (1956)
taxonomy was written to determine how the students were construct-
ing knowledge between what they had been learning in their science
lessons in class and the classification activity they had discussed in
their small groups when they had been videotaped. The questions on
the probe were formulated to ensure that the items had content valid-
ity for all the groups across the different schools. For example, a basic
recall question required the children to complete a close activity using
the words ‘characteristics, groups, classification, plants’. These were
words the students were familiar with from their work in classifying
living things. A more difficult activity required them to study a given
classification tree and to answer specific questions about the different
categories and sub-categories. At a more advanced level, the students
constructed their own key to classifying groups that they had pre-
viously identified and finally, to explaining why their key was useful.
This latter requirement was more difficult because it required the
students to synthesize information from a range of sources, evaluate its
usefulness and then justify their response.
Students were assigned a score from 1 (basic recall of fact/s) to 6
(evaluative response), depending on the highest level response they
were able to generate that was correct. For example, if a student was
able to answer a question that required him/her to apply information
correctly but he/she was unable to answer questions at a higher or
more difficult level, then a score of 3 (indicating the third level of

337

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

response) was assigned. The students’ responses to the questions on


the probe were marked by a science teacher and checked by a second
teacher with expertise in that area (neither teacher was involved
in teaching the students). Agreement between the teachers was 100
percent.

Procedure
The researcher met with the schools to discuss the preliminary assess-
ment involved in assigning students to groups, the procedures for
establishing groups (including the importance of ensuring that the
students knew they were to work together to assist each others’ learn-
ing on the activities assigned to the group) and the identification of
teachers who were willing to implement small-group work in Science
for a period of four to six weeks each term across the school year. All
schools received information on cooperative small group learning and
how it could be established in classrooms to promote learning.
The students were videotaped in their groups working on the group
problem-solving task (outlined above). The videotaping session
occurred during their second unit of science (four- to six-week period) to
allow students to adjust to the small group activities in their classroom.
Videotaping occurred in the last two weeks of the second Science unit to
ensure the students had covered the curriculum on which the activities
were based (i.e. understanding the principles of classification and how
it is applied to different life forms). The science probe which was
designed to ascertain how the students were constructing understand-
ings and making links between information discussed in the groups
was administered one to two weeks following the group videotaping
session.

Unit of analysis
The unit of analysis was the group rather than the individual because
it has been argued that it is important to acknowledge the non-
independence of individual members in cooperative groups (Kenny et
al., 2002). In this study, this included the students’ behaviours and dis-
course as they worked on the group problem-solving task together.

Results
To determine if there were significant differences in the Behaviour
States of the structured and unstructured conditions, a multivariate
analysis of variance (MANOVA) was conducted. Assumptions of nor-
mality and homogeneity of variance underpinning the use of MANOVA
were investigated. Normality was assessed by examining the variables
for skewness and kurtosis and while a few of the variable values were

338

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

Table 1 Means and SD of the frequency of Behaviour States for the


students in the Structured and Unstructured conditions
Structured Unstructured
condition condition
(n = 21 groups) (n = 23 groups)
Variable M SD M SD F-value*

Cooperation 11.93 1.56 10.65 2.45 4.24*


Noncooperation 0.07 0.14 0.51 0.73 7.32*
Individual task-oriented 0.45 0.56 1.08 1.05 5.98*
Individual non-task 0.50 0.51 1.06 0.81 5.70*

*p < 0.05; df = 1/42.

> 1, the majority of variables met the normality assessment. Homo-


geneity of variance was assessed by using the F-Max test.
The MANOVA was significant, T2 = 0.36 F(4, 39) = 3.54, p < 0.05,
h2 = 0.26 permitting an examination of the univariate results. Four
univariate results were significant, Cooperation, F (1, 42) = 4.24, p <
005, h2 = 0.09, Noncooperation, F(1, 42) = 7.32, p < 005, h2 = 0.14,
Independent Task-Oriented, F(1, 42) = 5.98, p< 0.05, h2 =0.12 and
Independent Non-task, F(1, 42) = 5.70, p < 0.05, h2 = 0.12. An exam-
ination of Table 1 shows that the children in the structured condition
displayed more cooperative behaviours and less non-cooperative, indi-
vidual task-orientated (working by self) and individual non-task
behaviours than their peers in the unstructured condition. The results,
however, need to be interpreted in relation to the strength of the effect
size which when less than 0.20 is regarded as small while 0.50 and 0.80
are regarded as medium and large for a two-group experiment
(Shaughnessy and Zechmeister, 1994).
To determine if there were significant differences in the Verbal Inter-
actions of the structured and unstructured conditions, a MANOVA was
conducted. Once again, assumptions of normality and homogeneity of
variance underpinning the use of MANOVA were again investigated
and found to have been met (as discussed above).
The MANOVA was significant, T2 = 4.32, F (9, 43) = 16.32, p <
0.001, h2 = 0.81, permitting an examination of the univariate results.
Five univariate results were significant: Directions, F(1, 42) = 8.16, p
< 0.01, h2 =0.16, Unsolicited explanations, F(1, 42) = 22.85, p <
0.001, h2 = 0.35, Unsolicited terminal, F(1, 42) = 12.14, p< 0.01, h2 =
0.22, Solicited explanation, F(1, 42) = 41.12, p < 0.001, h2 = 0.49 and
Non-specific interactions, F(1, 42) = 4.21, p < 0.05, h2 = 0.09. An
examination of Table 2 indicates that the children in the structured

339

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

Table 2 Means and SD of the frequency of Verbal Interactions for the


students in the Structured and Unstructured conditions
Structured Unstructured
condition condition
(n = 21 groups) (n = 23 groups)
Variable M SD M SD F-value*

Directives 1.75 0.79 2.72 1.35 8.16**


Directions + Physical prompts 2.97 1.69 2.12 1.16 3.74
Unsolicited explanations 2.68 1.21 1.17 0.85 22.85***
Unsolicited terminal 7.23 3.83 3.55 3.16 12.14**
Positive Interruptions 3.23 1.04 2.97 1.42 0.46
Negative interruptions 0.41 0.89 0.38 1.05 0.01
Solicited explanations 1.66 0.50 0.71 0.47 41.12***
Solicited terminal 3.80 1.92 3.67 2.10 0.04
Nonspecific verbals 8.45 7.98 13.27 7.59 4.21

**p < 0.01; ***p < 0.001; df = 1/42.

condition provided more help in the form of unsolicited explanations,


unsolicited terminal responses and solicited explanations than the
children in the unstructured condition. In contrast, the children in the
unstructured condition gave more directions and engaged in more non-
specific interactions. While the effect sizes obtained were generally
small (i.e. < 0.20), medium effect sizes were obtained for unsolicited
and solicited explanations (0.35 and 0.49) indicating that the difference
in these help giving behaviours was associated with group condition.

Cognitive language strategies


A MANOVA was also conducted on the frequency of the cognitive
language strategies used by the students to determine whether there
were differences between the structured and unstructured conditions.
The MANOVA was significant, T2 = 1.38, F(6, 37) = 8.53, p < 0.001,
h2 = 0.58, permitting an examination of the univariate results. Two
univariate results were significant; repeats information, F(1, 42) =
13.05, p < 0. 01, h2 = 0.23 and evaluative comment, F(1, 42) = 4.99,
p < 0.05, h2 = 0.10. An examination of Table 3 shows that the stu-
dents in the unstructured condition used more low-level cognitive
strategies such as repeating information than their peers in the
structured condition who expressed more evaluative comments or
higher-level cognitive strategies. However, the interpretation placed
on these results is limited by the small effect sizes obtained for the two
significant variables, repeats information and evaluative comments.

340

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

Table 3 Means and SD of the cognitive language strategies for the


Structured and Unstructured conditions.
Structured Unstructured
condition condition
(n = 21 groups) (n = 23 groups)
Variable M SD M SD F-value*

Repeats information 0.38 0.34 1.61 1.53 13.05**


Unstructured idea 3.73 1.50 2.91 2.23 1.96
Concrete idea 0.77 0.43 1.44 1.80 2.77
Explanation with evidence 1.22 0.54 1.01 0.75 1.18
Generalizes information 0.46 0.34 0.44 0.74 0.01
Evaluative comment 0.52 0.31 0.27 0.40 4.99*

*p < 0.04; **p < 0.01; ***p < 0.001; df = 1/42.

Science probe questionnaire


To determine if there were differences in the learning outcome scores
on the Science probe between the structured and unstructured condi-
tions, a one-way analysis of variance (ANOVA) was performed. The
ANOVA was significant, F(1, 141) = 116.82, p < 0.001, h2 = 0.10. An
examination of the means of the learning outcomes scores (Structured
condition M = 3.47, SD = 1.36; Unstructured condition M = 2.36,
SD = 1.82) showed that the children in the structured condition
obtained a higher learning outcome score than their peers in the
unstructured condition, although the interpretation placed on this
result is limited because of the small effect size.

Discussion
The present study was an investigation of the effects of cooperative
learning on junior high school students’ behaviours, discourse and
learning as they worked on a science problem-solving activity in struc-
tured and unstructured groups. The results show that the students
who worked in structured, cooperating groups demonstrated more
cooperative behaviour and less individually-orientated behaviour and
off-task behaviour than the students in the unstructured groups. In
short, their behaviour was more on-task and group focused. They
actively listened to what others had to say, shared ideas and resources
and provided both solicited and unsolicited explanations to assist with
learning. Abram et al. (2002) found that this type of behaviour is an
important predictor of the group’s overall productivity and perform-
ance while off-task talk within a group has a deleterious effect on that
group’s overall performance.

341

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

Although it has been demonstrated that in tasks that require a set


procedure and specific solution such as computational and algorithmic
tasks, it is the explanations that children provide to each other as they
work together that are critical for learning (Webb, 1992), others have
argued that in the context of tasks that are more open and discovery
based (i.e. where there is no set solution), it is the task-related dis-
course that affects the learning that occurs (Cohen, 1994). In this
study, the group problem-solving task was a non-routine problem for
which there were no standard procedures to follow so that the students
had to interact with each other if they were to arrive at a solution.
Furthermore, the task was constructed so no single individual in the
group was able to solve the problem without input from others. In
effect, discussion was vital for productivity and, in this study, this
included providing both solicited and unsolicited explanations.
However, if students are to benefit from the explanatory help they
receive, the explanations must be timely, relevant to the recipients’
need for help, of sufficient detail to enable them to clarify any mis-
understandings and correct in any information they provide (Webb et
al., 1995). Gillies and Ashman (1996, 1998) found that children who
work cooperatively together are often more perceptive of the needs of
their fellow students and will provide unsolicited help, such as expla-
nations and other helping responses, when they perceive they are
necessary. In short, although students can receive both solicited and
unsolicited explanatory help from their peers, this help must be of suf-
ficient elaboration for them to benefit from it (Webb, 2002). Certainly,
the help the students provided to each other in the structured groups
was detailed and coincided either with specific requests for help or with
students’ perceptions of the need to help and support others and it was
this help that undoubtedly contributed to the learning that occurred.
Moreover, an examination of the student discourse showed that the
children in the structured groups made more evaluative statements
that demonstrated the use of critical and reflective analysis of different
issues. Statements such as ‘It seems that we need to ...’ and ‘If we look
at these and do this ... we can see how they go together (make a new
category)’ and questions such as ‘What else do we still need to do ...?’
and ‘How are we going?’ were indicative of the types of self-monitoring
or meta-cognitive awareness that the students demonstrated. These
statements and reflections challenged their peers’ perspectives and
contributed to the engagement of group members around the group
problem-solving task. In contrast, the students in the unstructured
groups engaged in making more repetitive comments (i.e. repeating
information); language that is generally regarded as less likely to
promote higher-order thinking and problem solving among students
and less likely to challenge their peers’ interests in the task.

342

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

Evaluative talk in groups contributes to more product-related talk


which is a significant predictor of learning and achievement of group
members (Abram et al., 2002). Furthermore, when students engage in
critical and constructive talk with each other, it serves as an effective
tool in enhancing their problem-solving skills, both collectively and
individually (Rojas-Drummod et al., 2003). High quality talk is critical
for challenging children’s perspectives, developing problem-solving
skills and engaging them in learning (Cohen et al., 2002). Moreover, it
is more likely to emerge in structured cooperative settings where stu-
dents have been taught group processing skills that require them to
critique and reflect on their own and the group’s progress than in
unstructured groups where these skills have not been taught or were
taught in a more ad-hoc way. In this study, the students in the struc-
tured groups had been taught the importance of using key questions to
critique their achievements and reflect on their progress and this
undoubtedly contributed to their use of evaluative talk.
In addition to examining the differences in how the students in the
structured and unstructured groups interacted with each other, this
study also sought to measure how they were building understandings
and constructing links between information presented during the
group activities. The science probe that was administered to the
students at the completion of their work unit was specifically designed
to determine how they were building these understandings and con-
structing these links; in effect, generating new understandings and
knowledge to enable them to find new ways of thinking about issues
and solving problems (King, 2002).
The results showed that the children in the structured groups
demonstrated more complex and higher order thinking in responding
to specific problem-solving questions than their peers in the unstruc-
tured groups. The frequency of task-related interactions, such as the
explanations the students gave each other, provided clarification and
information on specific concerns. This, coupled with their willingness
to listen to each other and generally provide assistance when they
perceived it was needed contributed to a group ethos of care and con-
cern for each other and responsibility for each other’s achievements
(Slavin, 1996). It is these behaviours and attitudes that not only build
cohesiveness in a group and promote pro-social norms among group
members but they also act to motivate groups to engage in more self-
critical and reflective talk in their group discussions. It is the quality of
these discussions that are believed to affect the final learning outcome
(Cohen et al., 2002; Slavin, 1996).

343

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

Limitations
In the present study there were four characteristics which may limit
the conclusions that can be drawn. First, data on students’ behaviours
and interactions as they worked in their groups were only collected at
one point in time – during the last two weeks of the second science
curriculum unit. Follow-up observations may have demonstrated vari-
ations in students’ behaviours and interaction patterns that could not
be detected from a single observation. Second, no observations were
made of the teachers to determine how they employed small group
teaching strategies such as cooperative learning in their classrooms.
Thirdly, the small effect sizes limits the interpretation placed on the
results obtained. Finally, teachers’ perceptions of the ‘worthwhileness’
of cooperative group work in their classrooms were not sought yet such
information may have provided insights into why or why not it was
used. This is particularly pertinent given that Race and Powell (2000)
found that teachers in the higher grades report less favourable atti-
tudes towards using non-traditional teaching approaches.

Conclusion
The results of this study show that when schools are committed to
using cooperative learning, they need to ensure that teachers are
trained in how to establish cooperative learning activities in their
curricula and students are provided with opportunities to participate
in these activities on a regular basis. When this happens, students are
likely to engage in more helping behaviours such as giving more elabo-
rated help and guided directions to assist understanding and to
demonstrate more complex thinking and problem-solving skills both in
the content of their language and follow-up learning.

References
Abram, P., Scarloss, B., Holthuis, N., Cohen, E., Lotan, R. and Schultz, S.
(2002) ‘The Use of Evaluation Criteria to Improve Academic Discussion in
Cooperative Groups’, Asia Pacific Journal of Education 22: 16–27.
Antil, L., Jenkins, J., Wayne, S. and Vadsay, P. (1998) ‘Cooperative Learning:
Conceptualizations, and the Relation Between Research and Practice’,
American Educational Research Journal 35: 419–54.
Baines, E., Blatchford, P. and Kutnick, P. (2003) ‘Changes in Grouping Prac-
tices Over Primary and Secondary School’, International Journal of
Educational Research 39: 9–34.
Bloom, B. (1956) Taxonomy of Educational Objectives. Handbook 1: Cognitive
Domain. London: Longman.
Cohen, E. G. (1994) ‘Restructuring the Classroom: Conditions for Productive
Small Groups’, Review of Educational Research 64: 1–35.
Cohen, E., Lotan, R., Abram, P., Scarloss, B. and Schultz, S. (2002) ‘Can
Groups Learn?’, Teachers College Record 104: 1045–68.

344

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

Cohen, E., Lotan, R. and Holthuis, N. (1997) ‘Organizing the Classroom for
Learning’, in E. Cohen and R. Lotan (eds) Working for Equity in Heteroge-
neous Classrooms: Sociological Theory in Practice, pp. 32–43. New York:
Teachers’ College Press.
Cowie, H. and Berdondini, L. (2001) ‘Children’s Reactions to Cooperative
Group Work: A Strategy for Enhancing Peer Relationships Among Bullies,
Victims, and Bystanders’, Learning and Instruction 11: 517–30.
Gillies, R. (2003) ‘The Behaviors, Interactions, and Perceptions of Junior High
School Students During Small-Group Learning’, Journal of Educational
Psychology 95: 137–47.
Gillies, R. and Ashman, A. (2000) ‘The Effects of Cooperative Learning on Chil-
dren with Learning Difficulties in the Lower Elementary School’, The
Journal of Special Education 34: 19–27.
Gillies, R. and Ashman, A. (1998) ‘Behavior and Interactions of Children in
Cooperative Groups in Lower and Middle Elementary Grades’, Journal of
Educational Psychology 90: 746–57.
Gillies, R. and Ashman, A. (1996) ‘Teaching Collaborative Skills to Primary
School Children in Classroom-Based Work Groups’, Learning and Instruc-
tion 6: 187–200.
Johnson, D.W. and Johnson, F. P. (2003) Joining Together: Group Theory and
Group Skills, 8th edn. Boston, MA: Allyn and Bacon.
Johnson, D. W. and Johnson, R. (1990) ‘Cooperative Learning and Achieve-
ment’, in S. Sharan (ed.) Cooperative Learning: Theory and Research,
pp. 23–37. New York: Praeger.
Johnson, D. W. and Johnson, R. (2002) ‘Learning Together and Alone:
Overview and Meta-Analysis’, Asia Pacific Journal of Education 22: 95–105.
Johnson, D. W. and Johnson, R. (2000) ‘Cooperative Learning, Values, and
Culturally Plural Classrooms’, in M. Leicester, C. Modgil and S. Modgil
(eds) Education, Culture and Values, pp. 15–28. London: Falmer Press.
Johnson, D.W., Johnson, R., Dudley, B., Mitchell, J. and Fredrickson, J. (1997)
‘The Impact of Conflict Resolution Training on Middle School Students’, The
Journal of Social Psychology 137: 11–21.
Kamps, D., Dugan, E., Leonard, B. and Daoust, P. (1994) ‘Enhanced Small
Group Interaction Using Choral Responding and Student Interaction for
Children with Autism and Developmental Disabilities’, American Journal of
Mental Retardation 99: 60–73.
Kenny, D., Manetti, L., Pierro, A., Livi, S. and Kashy, D. (2002) ‘The Statistical
Analysis of Data from Small Groups’, Journal of Personality and Social Psy-
chology 83: 126–37.
King, A. (2002) ‘Structuring Peer Interaction to Promote High-Level Cognitive
Processing’, Theory into Practice 41: 33–40.
King, A. (1999) ‘Discourse Patterns for Mediating Peer Learning’, in A.
O’Donnell and A. King (eds) Cognitive Perspectives on Peer Learning,
pp. 87–152. London: Lawrence Erlbaum.
Kohn, A. (1992) ‘Resistance to Cooperative Learning: Making Sense of its
Deletion and Dilution’, Journal of Education 174: 38–55.
Lopata, C., Miller, K. and Miller, R. (2003) ‘Survey of Actual and Preferred Use
of Cooperative Learning Among Exemplar Teachers’, The Journal of Educa-
tional Research 96: 232–41.
Lou, Y., Abrami, P., Spence, J., Poulsen, C., Chambers, B. and d’Apollonia, S.
(1996) ‘Within-Class Grouping: A Meta-Analysis’, Review of Educational
Research 66: 423–58.

345

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


School Psychology International (2008), Vol. 29(3)

McMaster, K. and Fuchs, D. (2002) ‘Effects of Cooperative Learning on the


Academic Achievement of Students with Learning Disabilities: An Update
of Tateyama-Sniezek’s Review’, Learning Disabilities Research and Practice
17: 107–17.
Mercer, N., Wegerif, R. and Dawes, L. (1999) ‘Children’s Talk and the Develop-
ment of Reasoning in the Classroom’, British Educational Research Journal
25: 95–112.
Palincsar, A. (1998) ‘Social Constructivist Perspectives on Teaching and Learn-
ing’, Annual Review of Psychology 49: 345–75.
Palincsar, A. and Herrenkohl, L. (2002) ‘Designing Collaborative Learning
Contexts’, Theory into Practice 41: 26–35.
Race, K. and Powell, K. (2000) ‘Assessing Student Perceptions of Classroom
Methods and Activities in the Context of Outcomes-Based Evaluation’,
Evaluation Review 24: 635–46.
Rojas-Drummod, S., Perez, V., Velez, M., Gomez, L., Mendoza, A. (2003) ‘Talk-
ing for Reasoning Among Mexican Primary School Students’, Learning and
Instruction 13: 653–70.
Rutherford, R., Mathur, S. and Quinn, M. (1998) ‘Promoting Social Communi-
cation Skills Through Cooperative Learning and Direct Instruction’,
Education and Treatment of Children 21: 231–33.
Shachar, H. and Sharan, S. (1994) ‘Talking, Relating and Achieving: Effects of
Cooperative Learning and Whole-Class Instruction’, Cognition and Instruc-
tion 12: 313–53.
Sharan, S., Shachar, H. and Levine, T. (1999) The Innovative School: Organi-
zation and Instruction. Westpoint, CT: Bergin and Garvey.
Sharan, S. and Shachar, H. (1988) Language and Learning in the Cooperative
Classroom. New York: Springer-Verlag.
Shaughnessy, J. and Zechmeister, E. (1994) Research Methods in Psychology.
New York: McGaw-Hill.
Slavin, R. (1996) ‘Research on Cooperative Learning and Achievement: What
We Know, What We Need to Know’, Contemporary Educational Psychology
21: 43–69.
Slavin, R. and Cooper, R. (1999) ‘Improving Intergroup Relations: Lessons
Learned from Cooperative Programs’, Journal of Special Issues 55: 647–63.
Stevahn, L., Johnson, D., Johnson, R., Oberle, K. and Wakl, L. (2000) ‘Effects
on High School Students of Conflict Resolution Training Integrated into
English Literature’, Child Development 71: 772–84.
Stevens, R. (2003) ‘Student Team Reading and Writing: A Cooperative Learn-
ing Approach to Middle School Literacy Instruction’, Educational Research
and Evaluation 9: 137–60.
Turner, J. and Patrick, H. (2004) ‘Motivational Influences on Student Partici-
pation in Classroom Learning Activities’, Teachers College Record 106:
1759–85.
Veermans, M., Lallimo, J. and and Hakkarainen, K. (2005) ‘Patterns of Guid-
ance in Inquiry Learning’, Journal of Interactive learning Research 16:
179–94.
Vygotsky, L. (1978) Mind in Society: The Development of Higher Psychological
Processes. Cambridge, MA: Harvard University Press.
Webb, N. (2002) ‘Productive Helping in Cooperative Groups’, Theory into
Practice 41: 13–20.
Webb, N. (1992) ‘Testing a Theoretical Model of Student Interaction and
Learning in Small Groups’, in R. Hertz-Lazarowitz and N. Miller (eds)

346

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015


Gillies: Cooperative Learning in Junior High School

Interaction in Cooperative Groups, pp. 102–19. Cambridge: Cambridge Uni-


versity Press.
Webb, N. (1985) ‘Student Interaction and Learning in Small Groups: A
Research Summary’, in R. Slavin, S. Sharon, S. Kagan, R. Hertz-
Lazarowitz, C. Webb and R. Schmuck (eds) Learning to Cooperate,
Cooperating to Learn, pp. 5–15. New York: Plenum.
Webb, N., Baxter, G. and Thompson, L. (1997) ‘Teachers’ Grouping Practices in
Fifth-Grade Science Classrooms’, The Elementary School Journal 98:
91–113.
Webb, N. and Farivar, S. (1999) ‘Developing Productive Group Interaction in
Middle School Mathematics’, in A. O’Donnell and A. King (eds) Cognitive
Perspectives on Peer Learning, pp. 117–49. London: Lawrence Erlbaum.
Webb, N., Troper, J. and Fall, R. (1995) ‘Constructive Activity and Learning in
Collaborative Small Groups’, Journal of Educational Psychology 87:
406–23.

Robyn Gillies, PhD, is a Professor in the School of Education, The University


of Queensland, Brisbane, Australia. Her research focuses on peer-mediated
learning and small group processes. She is a trained teacher and psychologist.
Address: School of Education, The University of Brisbane, Australia, 4072.
Email: r.gillies@uq.edu.au

347

Downloaded from spi.sagepub.com at TULANE UNIV on January 23, 2015

You might also like