What email address or phone number would you like to use to sign in to Docs.com?
If you already have an account that you use with Office or other Microsoft services, enter it here.
Or sign in with:
Signing in allows you to download and like content, and it provides the authors analytical data about your interactions with their content.
Embed code for: Gilles, R. M. (2004) The effects of communication training on teachers and
Select a size
International Journal of Educational Research 41 (2004) 257–279 The effects of communication training on teachers’ and students’ verbal behaviours during cooperative learning Robyn M. Gillies School of Education, The University of Queensland, Brisbane 4072, Australia Abstract The present study sought to compare the effects of training teachers in speciﬁc communication skills designed to promote thinking and scaffold learning on teachers’ and students’ verbal behaviours during cooperative group work. Thirty teachers and 826 children from years 5 to 7 participated in the study. The results show that when teachers are trained to use speciﬁc communication skills during cooperative learning (cooperative-interactional condition) they engage in more mediated-learning interactions, ask more questions, and make fewer disciplinary comments than teachers who have been trained to implement cooperative group work only (cooperative condition). In turn, the children in the cooperative-interactional groups modelled many of the responses they gave their teachers and provided more detailed explanations, shorter responses, and asked more questions than their peers in the cooperative only groups. r 2005 Elsevier Ltd. All rights reserved. 1. Introduction Teachers play a critical role in promoting interactions among students and engaging them in the learning process. Cooperative, small-group learning is widely accepted as one way in which teachers can promote this interaction to beneﬁt all ARTICLE IN PRESS www.elsevier.com/locate/ijedures 0883-0355/$-see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijer.2005.07.004 E-mail address: email@example.com. students. When students interact in cooperative groups, they learn to give and receive information, develop new understandings and perspectives, and commu- nicate in a socially acceptable manner. It is through interacting with each other in reciprocal dialogues that children learn to use language differently to explain new experiences and new realities and, in so doing, construct new ways of thinking and feeling (Barnes, 1969; Mercer, 1996). The dialogues that occur are multi-directional as students learn to respond both to explicit and implicit requests for help and scaffold their responses to facilitate understanding and learning in others (Gillies & Ashman, 1996, 1998). Cooperative learning creates opportunities for students to actively interact with others, negotiate meaning around a task, and appropriate new ways of thinking and doing (King, 1999; Rogoff & Toma, 1997). By establishing a learning environment where children feel safe to test out their ideas, free from the scrutiny of the classroom teacher and the wider class group, they are provided with opportunities to reach out to each other and establish a personal synergy that facilitates engagement, promotion of learning, and group cohesion—all necessary elements for successful cooperative learning (Johnson & Johnson, 1990; Slavin, 1995). The beneﬁts of cooperative learning are well documented (Cohen, 1994; Johnson & Johnson, 2002; Slavin, 1995). However, much of this research has focused on how children learn in groups with little attention directed at the role teachers play in facilitating the learning that occurs. This may have happened because teachers have traditionally been encouraged to act as the guide on the side so children are expected to use each other as a resource rather than relying on direct help from their teacher (Hertz-Lazarowitz & Shachar, 1990). Moreover, this approach has been supported by an extensive volume of research that has documented the beneﬁts that accrue to children from interacting with others (e.g. King, 2002; Slavin, 1995; Webb, 1985, 1992). Certainly, social interaction plays a major role in how children learn in small group settings. However, teacher discourse and the role it may play in promoting learning during cooperative, small group learning has been less well documented. This is somewhat disconcerting given that it is the class teacher who often models how to engage in speciﬁc questioning techniques, scripted dialogues, reciprocal teaching, and exploratory talk—strategies that are used to help children talk and reason effectively together (see King, 1999; Mercer, Wegerif, & Dawes, 1999; O’Donnell, 1999; Palinscar & Brown, 1988; Rojas-Drummond & Mercer, 2003; Wegerif, Mercer, & Dawes, 1999). 1.1. Teacher verbal behaviours during cooperative learning While there is a paucity of research that focuses speciﬁcally on teachers’ verbal behaviours during small group learning, an early study by Hertz-Lazarowitz and Shachar (1990) provided some insight into the different discourses teachers use when they implement whole-class or cooperative, small-group learning in their classrooms. The study involved 27 elementary teachers who were trained to use cooperative learning as part of their repertoire of pedagogical practices. The results showed that when teachers implement cooperative learning in contrast to whole class instruction, ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279258 they increase their positive instructional behaviours (i.e. encouraging students’ efforts and facilitating student discussion) and drastically reduce their negative instructional behaviours such as disciplining students, interrupting students’ verbalisations and hurrying students when they work. In fact, the difference in the discourses was so marked that the authors concluded that when teachers implement cooperative learning, they become involved in a complex process of linguistic change as well so their language becomes more personal and intimate and less authoritian as they reach out to their students. In a study that built on this research, Gillies (in press) investigated whether teachers who implement cooperative learning in their classrooms engage in more facilitative, learning interactions with their students than teachers who implement small group-work only (i.e. the groups were not structured for cooperative learning). It has been argued that small-group work has many of the characteristics of the traditional, whole-class setting where there is no goal interdependence (i.e. a requirement that children work together to achieve a learning goal) and children work individually on tasks for their own ends. Hence, there is no motivation to act as a group or to exercise joint efﬁcacy to solve a problem or accomplish a task (Johnson & Johnson, 2003; Sharan, Shachar, & Levine, 1999). This study also investigated the verbal behaviours of the children in the cooperative groups to see if they model their teachers’ language behaviour and engage in more positive helping behaviours than their peers in the small group-work only condition. Given that teachers often model ways of talking to facilitate problem solving and reasoning, it was thought likely that the children may also seek to model these discourses as they interacted with each other. The study involved 26 teachers and 303 students in grades 8–10 from four large high schools in Brisbane, Australia. The results showed that teachers who implemented cooperative learning recorded more mediated-learning interactions (i.e. interactions designed to foster learning) and made fewer disciplinary comments than teachers who implemented small group- work only. Moreover, when teachers used these mediated-learning interactions, their tone and manner was soft and friendly and more personal and intimate as they worked with the children in their groups. Interestingly, a similar pattern of helping interactions was observed among the student groups with the children in the cooperative groups recording nearly twice as many elaborative responses designed to assist understanding and they were more verbally interactive than their peers in the group-work only condition. Both these verbal behaviours are important for fostering understanding and learning among group members (Cohen, 1994; King, 2002; Webb & Farivar, 1999). Given that the research indicates that teachers who establish cooperative learning in their classrooms engage in more facilitative verbal behaviour than teachers who teach in traditional, whole-class settings or teachers who implement small group- work only, the present study sought to determine if teachers could be trained to use speciﬁc communication skills to facilitate teacher and student discourses during cooperative learning. While there is a plethora of research that demonstrates that children can be taught speciﬁc cognitive and meta-cognitive skills to enhance thinking, understanding, and learning during peer-mediated instruction (see Fuchs, ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 259 Fuchs, Hamlett, Karns, & Dutka, 1997; King, 1999; King, Stafﬁeri, & Adelgais, 1998; O’Donnell, 1999; Palincsar, 1998), there is less information available on how these skills are employed by teachers in cooperative learning situations. Certainly, the role of high-quality interactions (e.g. explanations) in promoting learning in cooperative groups is widely acknowledged (Webb, 1992; Webb & Farivar, 1999). However, Meloth and Deering (1999) maintain that high-level talk (i.e. cognitively sophisticated talk) only emerges with low frequency when left to emerge naturally or as a by-product of cooperative group structures. Similarly, Chinn, O’Donnell, and Jinks (2000) found that children only used high-quality discourse with each other (even after training) when they were required to discuss reasons for their conclusions; and Mercer et al. (1999) found that children needed to be taught the ground rules necessary for engaging in exploratory talk if they were to use language effectively to reason and problem solve in their groups. In effect, Melroth and Deering, Chinn et al. and Mercer et al. believe that teachers need to be explicit in the skills they teach to enhance talk and children need to be encouraged to use these skills if they are to enrich discourse and enhance learning (Table 1). Schultz et al. (2000) found that when teachers used speciﬁc evaluation criteria about how students were to engage in subject matter and integrate different intellectual skills into the task, they focused on more academic-content talk with their students, were more speciﬁc in the feedback they gave, and were more likely to elicit evaluative comments from students than teachers who were not given the evaluation criteria. The evaluation criteria changed the way the teachers talked. While this study did not examine the effects of the teacher-talk on student interactions or learning outcomes, it does provide some interesting insights into the potential for teachers’ discourse to mediate students’ interactions and learning in cooperative groups. Certainly, given the Vygotsky (1978) perspective that more capable peers or adults scaffold or mediate children’s learning through social ARTICLE IN PRESS Table 1 Types and examples of communication skills teachers were trained to use Types of communication skills Examples of speciﬁc communication skills Probing and clarifying Can you tell me a little more about what you’re intending to do here? It seems as if you’re tryingy Perhaps you could tell us more about what you’re trying to say? Acknowledging and validating I can see you’ve worked really hard to ﬁnd out how these items are related. I wonder what you could do now to identify a key category they can all ﬁt into? Confronting discrepancies and clarifying options I wonder how you can include ...when you’ve already mentionedy? You’re telling me you can’t solve the problem, yet I notice you’ve managed to work out how these different parts can be used. It seems you might be part of the way there? Tentatively offering suggestions I wonder if you’ve considered doing it this way? Have you yy? Perhaps you could have a go at this and see if it’s suitable? R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279260 discourse, it seems logical to assume that teachers are also likely to engage in such verbal behaviours when they are involved in more intimate small group learning situations with their students. The present study builds on research that indicates that when teachers implement cooperative learning in their classrooms, it changes the way they interact with their students by seeking to determine if teachers can also be trained to use speciﬁc communication skills to facilitate student thinking and learning during cooperative group work. Two cohorts of teachers who were trained to embed cooperative learning strategies into their classroom teaching curricula participated in the study. One cohort of teachers received additional training in speciﬁc communication skills (cooperative-interactional condition) designed to facilitate teacher–student interac- tions during cooperative learning. The aims of the present study were to: (a) compare the effects of the cooperative-interactional and the cooperative conditions on teachers’ verbal behaviours during cooperative small group work; (b) compare the verbal behaviours of the student groups in the two conditions; and ( c) assess the learning outcomes of the students in the two conditions. 2. Method 2.1. Participants Thirty teachers from 11 primary schools in Brisbane, Australia volunteered to participate in the study. Schools were randomly allocated to either the cooperative- interactional (i.e. six schools) or cooperative condition (i.e. ﬁve schools) so that all teachers in the same school were in the same condition (This was done to ensure that there was less chance of the data being affected by teachers being in different conditions in the same school.). Four of the teachers were male and 26 were female. This broadly represents the ratio of male to female teachers (1:4) in Australian schools. All participating teachers were highly regarded by their respective school principals for their classroom management skills, their commitment to professional development, and their willingness to implement new ideas in their classrooms. The majority of the teachers were experienced teachers with 18 having taught for more than 11 years, six had taught for 6–10 years, and three had taught for 1–5 years and three did not indicate their length of teaching experience although all three had been teachers in their respective schools for longer than 3 years. In order to determine if there were differences between the teachers in the two conditions in their length of experience as teachers, a Mann Whitney U-test was performed because the data was ordinal and categorical (teachers were asked to indicate their length of service as follows: 1–5 years ¼ category 1, 6–10 years ¼ category 2, 11 years of more ¼ cate- gory 3). The result was not signiﬁcant (U ¼ 89:00, n1 ¼ 16, n2 ¼ 14, p ¼ 0:35, two- tailed) indicating that there were no differences between teachers in the two conditions in teaching experience. While 30 teachers participated in the study, complete data were only available on 28 teachers because of teacher absence on the day that the research team visited the ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 261 school (i.e. teacher illness). Teacher absence also affected the audio-taping that occurred in classrooms so that student groups were not audio-taped when their teachers were absent for their lessons (discussed below). Before the study began, all teachers were asked to complete the Teachers’ Orientation to Learning Questionnaire, an eight-item questionnaire that focused on their beliefs about teaching; for example, ‘‘I believe students learn best when they work with a partner, I believe students learn best when they plan their own work’’. The questionnaire tapped a number of core beliefs that teachers have about how student learn best from exercising more independence in learning through to being more reliant on the teacher’s directions and was informed by the National Competency Framework for Beginning Teachers (Australian Teaching Council, 1996). Teachers were asked to rate their responses from 1 (almost never) to 5 (almost always) to each of the statements about how they believe students learn best (Cronbach’s alpha ¼ 0.62). Table 2 presents the means and standard deviations of teachers’ responses to this questionnaire (see Table 2). A multivariate analysis of variance (MANOVA) showed that there were no differences between the teachers’ core beliefs about teaching for the teachers in the cooperative-interactional and cooperative conditions (discussed below), T2 ¼ 0:39, Fð8;21Þ¼1:04, p40:05, Z2 ¼ 0:28. 2.2. Students Eight hundred and twenty-six children in years 5–7 participated in the study (males ¼ 392, females ¼ 434; Mean age ¼ 10.91 years, SD ¼ 10:02 months). All the ARTICLE IN PRESS Table 2 Means and standard deviations of the teachers’ responses to the Teachers’ Orientation to Learning Questionnaire in the cooperative-interactional and cooperative conditions Variable Cooperative-interactional (n ¼ 16) Cooperative (n ¼ 14) M SD M SD 1. Learn best when working with a partner 3.81 0.75 3.64 0.84 2. Learn best when working in groups 3.52 0.62 3.42 0.85 3. Learn best when they have work autonomy 3.43 0.89 3.50 0.51 4. Learn best when working on activities they enjoy 4.56 0.51 4.78 0.42 5. Learn best when teachers have high expectations 4.68 0.47 4.62 0.49 6. Expectations set and grading known 4.43 0.62 4.64 0.63 7. Work to produce a group product 3.56 0.72 3.35 0.74 8. Students able to negotiate work for submission 3.81 0.83 4.21 0.80 NB: Ratings ranged from 1 ¼ almost never to 5 ¼ almost always. R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279262 schools in the study had a similar socio-economic and demographic proﬁle (i.e. 5–12% of the children came from different ethnic backgrounds; 10–15% of the children came from single-parent families; and over 75% of the parents were salaried workers). A one-way analysis of variance (ANOVA) of the stanine scores obtained by the children on the Otis–Lennon School Ability Test (OLSAT) (1993) indicated that there were no differences between the children in the two conditions (discussed below) (Fð1;506Þ ¼ 3:28, p40:05; cooperative-interactional M ¼ 5:20, SD ¼ 1:66; Cooperative M ¼ 4:95, SD ¼ 1:52). While 826 children participated in the study, complete data were only expected on 240 children because only eight children from each classroom were to be audio- taped. The ﬁnal data set from this latter group was 208 students. The incomplete data set for some students was due to absence from school either by the students or their teacher on the day that the research team visited the school. As discussed above, student groups were not audio-taped if their teacher was absent or if one of their members was absent. Stratiﬁed random assignment of students to their groups occurred within each class on the basis of their performance on the OLSAT (Otis & Lennon, 1993). The OLSAT is a group-administered test designed to measure verbal, quantitative, and ﬁgural skills that are closely related to school achievement. Students with low scores on the OLSAT usually have low scores on achievement measures, whereas those who excel in the ability measured by this test tend to obtain relatively high scores on different school subjects. This complex of abilities is assessed through performance on such tasks as detecting similarities and differences, solving analogies and matrixes, classifying, and determining sequence. The OLSAT comprises seven levels (Levels A–G) and two forms that collectively assess the range of abilities from kindergarten through to year 12. A scaled score system was developed for the OLSAT that links together all the levels (A–G) and both forms of the test, which makes it possible to compare the performance of students taking different levels or forms of the test. Reliability coefﬁcients for alternative forms range from 0.91 to 0.94 with an overall reliability coefﬁcient of 0.94. Raw scores were obtained by counting the number of correct answers and the total scores were converted into stanine scores. Stanine scores were used to reduce the likelihood of overestimation of small unreliable differences, as may occur in ﬁner scales such as IQ points. Each four-person group consisted of one high-ability student (top quartile of Otis–Lennon test), two medium-ability students (Quartiles 2 and 3), and one low-ability student (bottom quartile). Previous research has shown that while there is no evidence that one form of grouping is superior, low-ability children do beneﬁt from interacting with higher ability children and children of higher ability are not disadvantaged by working in mixed ability groups (Lou et al., 1996). However, teachers also used their discretion in assigning students to groups based on their knowledge of the children’s abilities to work cohesively together. In cases where difﬁculties arose, teachers re-assigned children to other groups or they established different grouping arrangements (i.e. two-member groups). ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 263 2.3. Cooperative-interactional and cooperative learning conditions All the teachers who participated in the study were volunteers and all indicated they were prepared to implement cooperative, small-group learning in their classrooms for one unit of work (i.e. 4–6 weeks) once a term for two terms (discussed below). The teachers participated in a 2-day workshop designed to introduce them to the basic tenets of cooperative learning and how to embed these tenets in their classroom curricula. The principal features of co-operative learning are as follows. Task interdependence exists when students perceive they are linked together in a way (e.g. task, goal) that they cannot succeed unless they all do; interpersonal and small-group skills involves an emphasis on such skills as listening, sharing tasks; individual accountability includes ensuring that all group members contribute to the group; promotive interaction involves offering help and explaining ideas; and, small-group processing skills involve the group in monitoring their progress (Johnson & Johnson, 1990). The workshop was experientially based to enable the teachers to develop a conceptual understanding of the theoretical foundations of cooperative learning, gain new insights into how it could be implemented in their curricula, and have the opportunity of embedding it in structured practice lessons for themselves. The workshop provided the teachers with the opportunity of interacting with their peers and discussing issues concerning the implementation of cooperative learning in their classrooms, reﬂecting on the beneﬁts for themselves, and receiving on-going support from their colleagues as they contemplated implementing this approach in their classrooms. In addition to the information that was presented on how to embed cooperative learning into classroom lessons, one cohort of teachers (cooperative- interactional condition) received additional training (i.e. one half-day of training) in the speciﬁc communication skills that promote meaningful engagement with the group task (see Table 1). These skills are used widely in counselling and are designed to challenge children’s understandings and perspectives with the intention of helping them focus more clearly on the problem to be solved (Egan, 2002; Ivey, 2002). The teachers spent time rehearsing these skills in triads and verbally processing the role plays with each other. In addition, they were asked to think and talk about their reactions to the role play situations, their assessment of the quality of the skills they were practising, and their hypothesis about how the children would react to these speciﬁc communication skills (Duys & Hedstrom, 2000). The researcher who conducted the workshops also moved among the group members and encouraged them to think and talk about their reactions to the use of these communication skills and how they might be used in classroom settings to promote discussion and thinking among group members. While the teachers in the cooperative-interactional condition received the additional training in speciﬁc communication skills, the teachers in the cooperative condition spent the same length of time working with each other and the researcher to embed the cooperative learning skills into speciﬁc classroom lessons. These lessons were then discussed with the larger group (i.e. all teachers in the cooperative condition). The researcher worked closely with these teachers discussing curriculum ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279264 issues, resources needed for planning these lessons, and their evaluations of the process and outcomes of cooperative learning. 2.4. Group activities The teachers were given guidelines on how to establish cooperative group activities in their classrooms. These guidelines included: ensuring that the task was structured so that all group members were required to contribute (positive interdependence); each child was held accountable for their contributions to the group (no social loaﬁng); children were taught the interpersonal and small-group skills needed to promote cooperation; the children worked in small groups of four members in face to face situations; and, the children used group processing skills to evaluate the group’s progress after each group session (Johnson & Johnson, 1990). In addition, the teachers were asked to ensure that the cooperative group activities required the children to consider information they had previously learned, and through a series of probing questions, demonstrate that they were able to apply, analyse, synthesise, and evaluate solutions to the task (Bloom, 1956). Many of the activities were open and discovery based where there were no correct answers and the children were required to cooperate as they discussed how to proceed as a group and share information (Cohen, 1994). The cooperative group activities were embedded into a unit of work each term for two school terms. While teachers were able to determine the content area that they would use for the cooperative learning activities, most chose to embed them into their social sciences programs (however, one chose English and one chose science), possibly because many of the learning experiences involve children in problem-solving activities where there are no correct answers and the children have to work together to discuss potential solutions. All teachers used the Studies of Society and Environment Years 1–10 Syllabus (Queensland School Curriculum Council, 2000) as a guide to the units of work they chose to develop. This syllabus is not prescriptive but rather provides a framework for planning learning experiences and assessment tasks through which students have opportunities to demonstrate what they know and can do in years 1–10 in this key learning area. The syllabus incorporates the cross-curricular priorities of literacy, numeracy, lifeskills and a futures’ perspective so that any unit of work would has an integrated curricular approach. Thus, while teachers designed their own units of work in which they embedded cooperative learning, they all used the same principles of planning for the learning experiences they designed for the students. 2.5. Measures 2.5.1. Teachers’ verbal behaviours The observation schedule (i.e. protocol) of the teachers’ verbal behaviours was originally developed by Hertz-Lazarowitz and Shachar (1990) and modiﬁed for the purposes of this study based on preliminary trialing of the categories in elementary classrooms where teachers used cooperative learning. Four teachers who did not participate in this study were videotaped for 45min as they used cooperative learning ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 265 in their classrooms (a total of 3h of videotapes). Two coders, using the Hertz- Lazarowitz and Shachar coding schedule, were used to identify the types of verbal behaviours the teachers used. The original 20 types of teachers’ verbal behaviours Hertz-Lazarowitz and Shachar identiﬁed were extended to include open questions and language that was required to maintain the activity. The 22 verbal behaviours were then collapsed into six categories of teachers’ verbal behaviours, ﬁve of which Hertz-Lazarowitz and Shachar had identiﬁed as representing the types of language that teachers use during cooperative learning in their classrooms (These categories emerged from a factor analysis of teacher verbal behaviours during cooperative and whole class learning reported by Hertz-Lazarowitz and Shachar and included teacher instructing/controlling, questioning, disciplining, helping, and encouraging.). The sixth category, mediates, was speciﬁcally developed for this study, although it was originally recognised as a teacher verbal behaviour by Hertz-Lazarowitz and Shachar Two additional coders who were experienced at coding language interactions from videotapes coded the original four teachers’ videotapes using the six categories of teachers’ verbal behaviours discussed above. Inter-observer agreement on each category of behaviour observed ranged from 85% to 100%. The categories of teachers’ verbal behaviours identiﬁed are presented in Table 3 (See Table 3). Verbal behaviours were coded according to frequency across each recorded class session and represent the total sum of statements per teacher (i.e. 100% of teacher talk) during that class period. The frequency of each category in each of the two recording periods was divided by the total sum of statements so that the percentage frequency score of each category, calculated separately, served as the unit of statistical analysis. This coding procedure was also used by Hertz-Lazarowitz and Shachar. A total of 42h of teachers’ verbal behaviour (i.e. 28 teachers were taped twice for a period of 45min) across the two time periods was collected using audio- microphones (discussed below). Three coders, blind to the purposes of the study, ARTICLE IN PRESS Table 3 Categories of teachers verbal behaviours Teacher control Instructing, lecturing, providing mechanical reinforcement to students, and reinforcement expressing comparisons between the children’s performances, initiative, or behaviours Questions Short questions and open questions which were designed to elicit expected information such as a short, unelaborated response Disciplines Discipline comments directed at individual students, groups, or the whole class Mediates Paraphrases to assist understanding, prompts, uses open questions in a tentative manner to promote thought about an issue the student is focused on, mediates or scaffolds learning between students to encourage engagement about an issue Encourages Praises student’s, group’s, or class efforts, encourages interactions among students, and expresses spontaneous emotion (e.g. ‘‘I’m really excited with the way this group are handling their task’’) Maintenance Helps the child during learning, refers to the problem task without punishing, refers to technical problems in carrying out the task, and all language needed to maintain the activity. R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279266 coded a common 3h of audiotape and inter-rater reliability ranged from 92% to 100% for the categories coded. 2.5.2. Students’ verbal behaviours The observation schedule for the students’ verbal behaviours was based on a schedule developed by Webb (1985), adapted by Gillies (2003), and modiﬁed for the purposes of this study to gather information on the types of verbal behaviours that the students used during cooperative learning. Modiﬁcations to this schedule occurred after trails conducted in four elementary classes (same classes in which teachers’ verbal behaviours were videotaped as discussed above) where children’s verbal behaviours during cooperative learning were videotaped and analysed. Two coders, using the Gillies (2003) coding schedule, coded 3h of videotape (4 lessons45min each; one group each lesson). The original 13 verbal behaviour variables identiﬁed by Gillies were grouped into six categories and two additional coders used these categories to code the original 3h of audiotape. Inter-observer agreement on each category of verbal behaviour identiﬁed ranged from 80% to 98%.The categories of verbal behaviour identiﬁed that students use during cooperative learning are presented in Table 4 (See Table 4). Student verbal behaviours were coded according to frequency of statement across each recorded session and represent the total sum of group verbal behaviours (i.e. 100% of group talk) during that class period. The frequency of each category in each of the two recording periods was divided by the total sum of verbal behaviours so that the percentage frequency score of each category (calculated separately) served as the unit of statistical analysis. A total of 78h of students’ verbal behaviours during cooperative learning was collected (i.e. 52 groups were audio-taped twice for a period of 45min each time). Once again, the same three raters coded 3h of the students’ verbal behaviours and inter-rater reliability ranged from 89% to 100% for the categories coded. 2.5.3. Learning outcome questionnaire Teachers were provided with a model learning outcomes questionnaire which they used to enable them to gauge the extent to which the children were building ARTICLE IN PRESS Table 4 Categories of students’ verbal behaviours Elaborations Provides detailed explanations, expands on another student’s comments Questions Open and closed questions Short responses Unelaborated responses consisting on ‘yes’ or ‘no’ or a short phrase designed not to encourage further interaction Engages Acknowledges another student’s effort, engages in a sustained exchange on the other student’s comment/s, positive interruption to contribute to the discussion. Directs Gives instructions for another student to follow, disciplines another student to focus attention Interrupts Negative interruption designed to detract from the discussion or pass a derogatory comment R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 267 understandings and making connections between information presented and discussed during their cooperative learning experiences. The learning questionnaire consisted of a set of generic questions that were originally developed by King (1991, 1994) and modiﬁed by Gillies and Ashman (1996, 1998) and Gillies (2003). Questions were arranged according to the level of complexity or depth of thinking required. For example, less complex questions asked children to recall facts or basic details. Such recall questions usually began with one of the following sentence stems: What is y ? Name as many y Questions of increasing complexity followed. For example, comprehension questions began with: What do you think y? Explain why y Application questions began with: Examine the y; Using the facts you know y More complex questions asked children to analyse and integrate information or evaluate difference sources of information to propose a solution or reason to solve a problem. For example, questions that required children to analyse information began with one of the following sentence stems: Comparey with y. Examine the y and discuss their attributes. How might you categorise them? Questions that encouraged children to synthesise information began with: Imagine you are a y Find another way to y Finally, questions that required children to evaluate information began with: Select and justify y; Discuss the pros and cons of y Each teacher built their own content on to the learning outcome questionnaire which was checked by two teachers in the research team and the author to ensure it was a valid measure of the unit of work the children had completed in their cooperative groups (i.e. that the questionnaire covered equivalent content and an appropriate range of material within the unit of work and accurately reﬂected all six levels of Bloom’s taxonomy). The questionnaire was marked by one of these teachers and checked by the second teacher and the author. Inter-rater reliability for the marking of the questionnaires was 100%. The classroom teachers did not mark these questionnaires. Children were assigned a learning outcome score of 1–6 depending on the highest level response they were able to complete correctly. For example, if a child was able to answer a question that required him/her to apply information to a situation/ problem but was unable to analyse information indicating the inter-relationships of the component parts, then a learning outcome score of 3 (indicating the third level of response) was assigned. On the other hand, if a child was able to correctly analyse information but was unable to apply it to a situation/problem, then he/she was assigned a learning outcome score of 4 even though he/she missed the lower level response. The advantage of using this type of measure was that it allowed teachers, by using the generic questions provided, to construct a learning outcome measure that was authentic and relevant to the unit of work the children had completed. 2.5.4. Teacher application of a cooperative learning framework The observation schedule for the teacher application of a cooperative learning framework was developed speciﬁcally for this study but was informed by the key elements of cooperative learning (Johnson & Johnson, 1990) discussed above. The ﬁve dimensions that were observed were: (a) uses a range of cooperative learning strategies and orchestrates their use according to the stage of curriculum ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279268 development (e.g. pair-share, round robin, reciprocal questioning, brainstorming among the group—activities which were designed to encourage student interaction and discussion), (b) uses language that reﬂects the fact that cooperative learning strategies are being employed (i.e. responsibilities for tasks, roles), (c) facilitates the students’ use of cooperative learning (i.e. encourages students to work together, reduces explicit teaching and encourages group members to seek help from each other), (d) reinforces the students’ use of learning strategies (e.g. use of encouragement, use of praise, reﬂection sheets for group processes and tasks), and (e) develops interdependence in the students (e.g. shared response, group chart to share group response, attention to grouping students). Three research assistants (all teachers who had been trained speciﬁcally in the behaviours to observe) rated each dimension using a Likert scale from 1 to 5 to indicate whether the behaviour was not observed at all (1), to whether it was observed almost always (5). Finally, an overall rating of 1–5 was made on the implementation of a cooperative learning framework in the lesson being observed. The three observers each observed the same two lessons so that a total of six lesson were observed. Inter-observer reliability on the application of the cooperative learning framework ranged from 95% to 100% across the ﬁve dimensions rated. Inter-observer ratings on the overall rating was 100%. All teachers obtained ratings of 3–5 and were deemed to have implemented cooperative learning in their classrooms. The Mann Whitney U-test was performed on the overall rating of the implementation of the cooperative learning framework and the result was not signiﬁcant (U ¼ 85:00, N1 ¼ 14, N2 ¼ 14, p ¼ 0:57, two-tailed) indicating that there were no differences between the teachers in the two conditions in the way they implemented cooperative learning. 2.6. Procedure The researcher discussed the preliminary testing and the assignment of students to groups, and the data collection process with the teachers prior to the commencement of the study. During the 2-day workshop (discussed above), the teachers received additional information on the background to the research and the procedures for establishing cooperative learning in their classrooms. The teachers were audio-taped twice during lessons in which they used cooperative learning activities. The audio-taping occurred towards the end of the ﬁrst and ﬁnal unit of work in which they used cooperative learning strategies. The teachers wore an audio-microphone and they were taped for the full class period in which the children worked on their small-group activities. This covered a period of 45min. Samples of the children’s language from two of the small groups, chosen randomly, initially, from the groups in each classroom (the same groups were then audiotaped at both taping sessions), were collected during these taping sessions by placing a cassette recorder on the table for the duration of the small-group activity. In addition, an observer sat discreetly at the back of each classroom and completed an observation schedule on the implementation of the cooperative learning framework. A mixed sample of 10 children’s learning outcome questionnaire responses (3 high-achieving, 4-average achieving, & 3-low achieving) were collected for each class at the ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 269 completion of the second unit of work. This sample provided the measure for the learning outcomes for the children in the two conditions. 3. Results 3.1. Teachers’ verbal behaviour categories The means and standard deviations of the percentage of verbal behaviour categories for the teachers in the cooperative-interactional and cooperative conditions for Times 1 and 2 are presented in Table 5 (see Table 5). To determine if there were signiﬁcant differences in the verbal behaviour categories of the teachers in the two conditions across time, a Condition by Time MANOVA was conducted with a repeated measure on the last dimension. The MANOVA yielded a signiﬁcant multivariate effect for Condition, T2 ¼ 5:59, Fð6;21Þ¼19:58, po0:001, Z2 ¼ 0:84, but not for Time, T2 ¼ 0:23, Fð6;21Þ¼0:82, p ¼ 0:56, Z2 ¼ 0:19, or Time by Condition, T2 ¼ 0:22, Fð6;21Þ¼0:77, p ¼ 0:60, Z2 ¼ 0:18. Because a signiﬁcant effect was found for Condition, an examination of ARTICLE IN PRESS Table 5 Means and standard deviations of the categories of verbal behaviours of the teachers (represented in percentages) in the cooperative-interactional and cooperative conditions at Times 1 and 2 Condition Time Cooperative-interactional (n ¼ 14) Cooperative (n ¼ 14) 1 2 1 2 Control M 17.49 15.97 20.90 22.82 SD 6.98 5.44 6.62 5.75 Question M 38.25 39.70 23.10 21.20 SD 12.44 13.26 5.18 7.10 Discipline M 1.05 1.22 5.82 5.76 SD 1.57 1.83 3.48 2.35 Mediate M 14.75 12.02 7.79 7.61 SD 3.05 5.02 4.54 3.69 Encourage M 15.75 15.07 20.56 18.31 SD 6.38 6.11 5.72 4.58 Maintenance M 12.55 13.68 21.17 24.20 SD 5.98 3.73 6.65 5.11 R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279270 the univariate results was permitted. To protect against Type 1 error, the Bonferroni adjustment (Keppel, 1991) to the nominal alpha was undertaken before reporting all univariate results (NB: this adjustment was made to all univariate results on all MANOVAs in this study). Thus, the probability of all group comparisons was made equal to 0.05 (k1)/k where k is the number of tests in each comparison. Six univariate results were signiﬁcant, Control, Fð1;26Þ¼7:67, po0:04, Z2 ¼ 0:22, Question, Fð1;26Þ¼34:45, po0:001, Z2 ¼ 0:57, Discipline, Fð1;26Þ¼42:49, po0:001, Z2 ¼ 0:62, Mediate, Fð1;26Þ¼26:93, po0:001, Z2 ¼ 0:50, Encourage, Fð1;26Þ¼4:94, po0:04, Z2 ¼ 0:16 and Maintenance, Fð1;26Þ¼33:14, po0:001, Z2 ¼ 0:56. An examination of Table 5 shows that the teachers in the cooperative-interactional condition engaged in more questioning and mediated-learning behaviours than their peers in the cooperative condition. In contrast, the teachers in the cooperative condition engaged in more controlling, disciplining, and encouraging verbal behaviours. They also engaged in more maintenance-type verbal behaviours than their peers in the cooperative-interactional condition. 3.2. Students’ verbal behaviours To determine if there were signiﬁcant differences in the verbal behaviour categories of the student groups in the cooperative-interactional and cooperative conditions across time, a Condition by Time MANOVA was conducted with a repeated measure on the last dimension. The MANOVA yielded a signiﬁcant multivariate effect for Condition, T2 ¼ 4:97, Fð6;45Þ¼37:32, po0:001, Z2 ¼ 0:83, Time, T2 ¼ 0:41, Fð6;45Þ¼3:09, po0:05, Z2 ¼ 0:29, and for Time by Condition, T2 ¼ 0:44, Fð6;45Þ¼3:36, p ¼ o0:01, Z2 ¼ 0:31 permitting an examination of the univariate results. A summary of the univariate F-tests and effect sizes for Condition, Time, and Time by Condition are presented in Table 6 (see Table 6). An examination of Table 7 of the means and standard deviations of the students’ verbal behaviours shows that the children in the cooperative-interactional condition gave more elaborations, questions, and short responses while their peers in the cooperative condition gave more engagement, directions, and interruption ARTICLE IN PRESS Table 6 Summary of univariate F-tests and effect sizes for Condition, Time, and Time by Condition for the students’ verbal behaviours Variable Condition F Z2 Time F Z2 TimeCondition F Z2 Elaboration 14.98*** 0.23 11.47* 0.18 0.01 0.00 Question 15.42*** 0.23 0.04 0.00 1.90 0.00 Short response 111.90*** 0.69 7.36** 0.12 14.77*** 0.22 Engagement 95.39*** 0.65 0.37 0.00 4.99* 0.09 Directs 4.71* 0.08 87.86*** 0.63 3.07 0.06 Interrupts 3.80 0.07 0.23 0.00 0.07 0.09 df 1/50.*po0.04, **po0.01, ***po0.001. R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 271 responses. However, the students’ verbal behaviours in their groups were inﬂuenced not only by the condition they were in (i.e. cooperative-interactional or cooperative) but also by changes over time for elaborations, short responses, and directions and the interaction of time by condition for short responses and engagement. Hence, the children in the cooperative-interactional condition gave more short responses from Time 1 to Time 2 while the children in the cooperative condition gave fewer short responses across the same time period. In contrast, the children in the cooperative- interactional condition gave fewer engagement responses from Time 1 to Time 2 while their peers in the cooperative condition increased their engagement behaviours. 3.3. Learning outcomes questionnaire In order to determine if there were differences in the learning outcomes for the children in the cooperative-interactional condition and those in the cooperative condition, a one-way ANOVA was performed on the scores obtained on the learning outcome questionnaire that was administered at the completion of the second unit during which the children had worked cooperatively together. The ANOVA was ARTICLE IN PRESS Table 7 Means and standard deviations of the categories of students’ verbal behaviour (represented in percentages) in the cooperative-interactional and cooperative conditions at Times 1 and 2 Condition Time Cooperative-interactional (n ¼ 32) Cooperative (n ¼ 20) 1 2 1 2 Elaboration M 14.45 8.43 7.77 2.12 SD 9.73 9.48 8.05 2.90 Question M 20.63 23.38 16.73 14.69 SD 7.85 10.07 6.12 7.57 Short response M 24.87 38.32 11.79 9.47 SD 12.21 11.52 6.17 5.39 Engagement M 17.81 10.44 33.14 37.33 SD 10.56 6.68 10.79 19.09 Directs M 18.94 8.09 24.85 9.00 SD 9.57 5.16 8.98 4.12 Interrupts M 3.62 3.88 5.59 6.55 SD 6.16 5.49 5.92 6.79 R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279272 signiﬁcant, Fð1;220Þ¼4:28, po0:05, Z2 ¼ 0:02. An examination of the learning outcome scores for the two conditions (cooperative-interactional M ¼ 3:78; SD ¼ 1:49; cooperative M ¼ 3:43, SD ¼ 1:07) showed that, although the children in the cooperative-interactional condition attained signiﬁcantly higher scores than their peers in the cooperative condition, the effect size (Z2 ¼ 0:02) was very small, limiting its educational signiﬁcance (effect sizes of 0.22+ are regarded as educationally signiﬁcant). 4. Discussion The present study sought to: (a) compare the effects of the cooperative- interactional and the cooperative conditions on teachers’ verbal behaviours during cooperative group work; (b) compare the verbal behaviours of the students in the two conditions; and (c) assess the learning outcomes of the students. The study was conducted across two school terms and required the teachers to implement cooperative learning in one unit of work (4–6 weeks) each term. The results show that the teachers who had been taught speciﬁc communication skills to promote thinking and to scaffold children’s learning (cooperative-interactional condition) engaged in more verbal behaviours such as questioning and mediated-learning behaviours or verbal behaviours designed to facilitate meaningful engagement with the group task than teachers who had not been taught these speciﬁc skills. While the questioning behaviour used by these teachers was somewhat disappointing, given that the questions often elicited expected information such as short, unelaborated responses, Turner et al. (2002) argues that this type of verbal behaviour can be effective if used in combination with instructional scaffolding or, as occurred in this study, mediated-learning behaviours. These latter behaviours included tentatively offering suggestions, probes that were designed to encourage children to consider an alternative perspective, or acknowledgements of their efforts while focusing on key issues or solutions. These mediated-learning behaviours set up a sequence of reciprocal interactions between the teachers and students that helped to focus the children’s attention on speciﬁc problem-solving strategies and encouraged continu- ing engagement with the task. These skills are used widely in counselling to promote thinking and problem-solving in individuals (Egan, 2002; Ivey, 2002). However, their use by teachers to facilitate discussion and problem-solving during cooperative learning is not well documented. Certainly, teachers use a range of cognitive and metacognitive skills to promote thinking in children during small-group work (e.g. peer tutoring, peer-mediated learning, and cooperative learning) but the skills used tend to focus on asking speciﬁc questions (e.g. ‘tell me why’?) or learning speciﬁc scripts (e.g. reciprocal teaching, guided instruction) with the outcome measure usually being the effect these skills have on children’s learning. Little attention has focused on how training teachers to use speciﬁc communication skills affects the way they interact with their students during cooperative small group activities. This is particularly important given that Schultz et al. (2000) were able to demonstrate that when teachers use speciﬁc ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 273 evaluation criteria, it changed the way they talked about academic work. While the potential for teacher-talk to mediate student-talk in groups was not explored in this study, Gillies (in press) was able to demonstrate that when teachers engage in more facilitative learning and prosocial interactions, as they do when they implement cooperative learning, students model many of their teachers’ verbal behaviours and provide more detailed or elaborative help to other group members. Moreover, Cohen, Lotan, Abram, Scarloss, and Schultz (2002), in a study on the use of evaluation criteria to promote group discussion and learning, suggested that students probably learn new ways of thinking and talking by listening to teachers model these behaviours in their interactions with students. Furthermore, it is the quality of the talk that groups generate that is a signiﬁcant predictor of students’ learning (King, 2002; Webb & Farivar, 1999). The current study found that when teachers are trained to use speciﬁc communication skills during cooperative learning they engage in more mediated- learning behaviours, ask more questions, and make fewer disciplinary comments than teachers who have not been trained to use these skills. The following are examples of the types of mediated-learning comments that the one teacher made during her interactions with a small group in her classroom who were working on energy sources: ‘‘I notice you’re working on looking at the different types of energy use and I can see you’re looking at nuclear powery I wonder how you could determine what proportion of energy comes from this source?’’ (Validation of efforts and tentative question designed to focus the children’s attention on how they might seek the answer they need from the pie graph and other information sources they are using), ‘‘Can I just suggest one thing? (teacher seeks permission from the group to offer a suggestion)yYou’ll need to think about referencing your information so if someone challenges it, you’ve got your source.’’(focusing the group’s attention on the need to document their evidence), ‘‘So you could say give a short description of green house gasses then what could you do?’’ (teacher offers suggestion on what the group may need to do and then probes to see if they have thought about what could happen next). ‘‘Now can I askyhow does burning coal become electricity?’’ (teacher challenges the children to see if they can make the link in the stages of energy development). It may be that when teachers use these types of verbal behaviours, they model not only how to engage in problem-solving discourse or thinking about thinking (King, 2002) but also how to participate in developing shared understandings of the issue at hand (Palinscar & Herrenkohl, 2002). In effect, children learn how to respond to their teachers’ verbal behaviours and become more active in their own learning. Certainly, the importance of teachers modelling desired behaviours in their interactions with children is one of the ways to promote productive helping in groups (Webb, Farivar, & Mastergorge, 2002). In this case, however, the teachers modelled not only the thinking behaviours but also the behaviours that invite participation as partners in resolving the problem at hand. This is somewhat different to the instructional discourse used more widely by teachers where they ask questions, students respond, and teachers then evaluate the answer (Turner et al., 2002). ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279274 Given that the research on children’s helping behaviours indicates that children who work in cooperative groups provide more detailed explanations and elaborated responses to other group members than children who work in whole class settings or small groups only and it is these types of help that mediate the learning that occurs (Webb, 1992; Webb & Farivar, 1999), it could be expected that the children’ helping behaviours, across the two conditions, would be similar because they worked in cooperative groups. However, the children in the cooperative-interactional condition recorded more elaborations, questions, and short answer responses while the children in the cooperative condition recorded more engaging, directing and interrupting behaviours. Although the interpretation placed on these results is limited by the effect sizes obtained (see Table 4), it is worth noting that it was the children in the cooperative-interactional condition who demonstrated more of those verbal behaviours that are associated with productive helping such as providing elaborated and detailed assistance to others (Gillies, 2003; Webb & Mastergeorge, 2003). While this study did not investigate a direct link between the teachers’ verbal behaviours and those of the children (i.e. reciprocal interactions were not investigated), it is interesting to note that the children’s responses to each other mirrored many of the responses they gave their teachers, that is, they were detailed or elaborated (e.g. When we talk about green house effect, we need to talk about the burning of fossil fuels and their effects on the green house cause they warm up the climate. Student provides elaborated response) and focused (e.g. Do they all cause the green house effect? What else causes the green house effect? Student focuses the group’s attention on the topic and seeks clariﬁcation on whether all the ideas discussed are relevant). It appears that when teachers are explicit in the types of thinking they want children to engage in, it encourages the children to be more focused and explicit in the types of help they provide to others. In effect, teachers can structure interactions in groups that guide and support high-quality thinking and discussion (King, 2002). Furthermore, it is the quality of the discourse that emerges that enhances learning (Chinn et al., 2000). This certainly happened in this study although the small effect size obtained on the learning outcome measure limits the interpretations placed on this result. 4.1. Theoretical implications One important theoretical implication that emerged from the study is the role that the teachers play in the social construction of knowledge at both the inter- and intra- personal level. The mediated-learning interactions that the teachers used were designed to not only scaffold the children’s learning but also to prompt meaningful cognitive and metacognitive thinking about the problem-solving activities (King, 1999; Palincsar, 1998). Because the teachers’ language was both tentative and inviting, they were able to scaffold or mediate potential learning while, also, challenging the children’s thinking and encouraging them to consider alternative points of view (King, 2002). In doing so, they were not only introducing the children to new patterns of thought but also to new ways of learning how to reach consensus ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 275 about what they were thinking and how that information may be used to resolve the problem at hand (Palinscar & Herrenkohl, 2002). It has been argued that it is through repeated exchanges with adults or peers in mutually cooperative interactions that children’s own thinking becomes inﬂuenced by these communicative exchanges and they begin to adopt them as their own (Damon, 1984; Shachar & Sharan, 1994). The responses expected from this type of discourse appears to create an expectation in recipients to reconcile or justify their perspectives with those of others, clarify misunderstandings, and provide responses that others will accept as valid. This coupled with the intimacy of the small group may have provided a psychological environment that motivated the children to be more willing to reconcile contra- dictions between themselves and others, test-out their ideas, and work to construct new understandings (Johnson & Johnson, 2003). 4.2. Limitations These are several limitations to this study. Firstly, there were only two data collection points and this may have affected opportunities to see changes over time in the teachers’ and students’ verbal behaviours in the two conditions. Secondly, the intervention was only conducted in one subject area once a term for two terms so there was no opportunity to see how the differences that were recorded in students’ verbal behaviours may have generalised to the wider classroom curriculum. Finally, the study did not investigate the reciprocal interactions that occurred between the teachers and students as the teachers moved among the small groups and interacted with different group members. Had this been done, it would have provided a more comprehensive understanding of the nature of the reciprocity that occurred and how, speciﬁcally, different types of teacher-initiated behaviours contribute to the types of responses children generate. These are issues that future studies may need to address. 5. Conclusion The study shows that when teachers are trained in speciﬁc communication skills designed to promote thinking and scaffold learning during cooperative learning, they engage in, more mediated-learning interactions, ask more questions, and make fewer disciplinary comments than teachers who have been trained to establish cooperative learning only. In turn, the children in the cooperative-interactional groups modelled many of the responses they gave to their teachers and provided more detailed explanations, short-answer responses, and asked more questions than their peers in the cooperative groups. While the study did not draw direct links between the types of verbal behaviours and the learning that occurred, it was interesting to note that the children in the cooperative-interactional condition obtained higher learning outcome scores than their peers in the cooperative condition. It seems that children’s discourse is enriched and learning enhanced when teachers are taught how to engage in problem-solving interactions with their students. These ﬁndings provide support ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279276 for the important role teachers can play in enhancing children’s discussions during cooperative learning. References Australian Teaching Council. (1996). National competency framework for beginning teachers. National project on the quality of teaching and learning. Canberra: Australian Government Publishing Service. Barnes, D. (1969). Language in the secondary classroom. In D. Barnes, J. Britton, & H. Rosen (Eds.), Language, the learner, and the school (pp. 11–76). Harmonsworth, Middlesex, England: Penguin Books. Bloom, B. (1956). Taxonomy of educational objectives. Handbook 1: Cognitive domain. London: Longman. Chinn, C., O’Donnell, A., & Jinks, T. (2000). The structure of discourse in collaborative learning. The Journal of Experimental Education, 69, 77–89. Cohen, E. (1994). Restructuring the classroom: Conditions for productive small groups. Review of Educational Research, 64, 1–35. Cohen, E., Lotan, R., Abram, P., Scarloss, B., & Schultz, S. (2002). Can groups learn? Teachers’ College Record, 104, 1045–1068. Damon, W. (1984). Peer education: The untapped potential. Journal of Applied Developmental Psychology, 5, 331–343. Duys, D., & Hedstrom, S. (2000). Basic counselor skills training and counselor cognitive complexity. Counselor Education and Supervision, 40, 8–18. Egan, G. (2002). The skilled helper: A problem-management and opportunity-development approach to helping (7th ed). Paciﬁc Grove, CA: Brooks/Cole. Fuchs, L., Fuchs, D., Hamlett, C., Karns, K., & Dutka, S. (1997). Enhancing student helping behavior during peer-mediated instruction with conceptual mathematical explanations. The Elementary School Journal, 97, 223–249. Gillies, R. (in press). Teachers’ and students’ verbal behaviours during cooperative and small-group learning. British Journal of Educational Psychology, in press. Gillies, R. (2003). The behaviors, interactions, and perceptions of junior high school students during small- group learning. Journal of Educational Psychology, 95, 137–147. Gillies, R., & Ashman, A. (1996). Teaching collaborative skills to primary school children in classroom- based work groups. Learning and Instruction, 6, 187–200. Gillies, R., & Ashman, A. (1998). Behavior and interactions of children in cooperative groups in lower and middle elementary grades. Journal of Educational Psychology, 90, 746–757. Hertz-Lazarowitz, R., & Shachar, H. (1990). Teachers verbal behavior in cooperative and whole-class instruction. In S. Sharan (Ed.), Cooperative learning: Theory and research (pp. 77–94). New York: Praeger. Ivey, A. (2002). Intentional interviewing and counseling: Facilitating client development in a multicultural society. New York: Wadsworth. Johnson, D. W., & Johnson, R. T. (1990). Cooperative learning and achievement. In S. Sharan (Ed.), Cooperative learning: Theory and research (pp. 173–202). New York: Praeger. Johnson, D. W., & Johnson, R. T. (2002). Learning together and alone: Overview and meta-analysis. Asia Paciﬁc Journal of Education, 22, 95–105. Johnson, D. W., & Johnson, R. T. (2003). Student motivation in cooperative groups: Social interdependence theory. In R. M. Gillies, & A. F. Ashman (Eds.), Cooperative learning: The social and intellectual outcomes of learning in groups (pp. 136–176). London: RoutledgeFalmer. Keppel, G. (1991). Design and analysis (3rd ed). New York: Prentice-Hall. King, A. (1991). Improving lecture comprehension: Effects of a metacognitive strategy. Applied Cognitive Psychology, 5, 331–346. King, A. (1994). Guided knowledge construction in the classroom: Effects of teaching children how to question and how to explain. American Educational Research Journal, 31, 338–368. ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 277 King, A. (1999). Discourse patterns for mediating peer learning. In A. O’Donnell, & A. King (Eds.), Cognitive perspectives on peer learning (pp. 87–115). Mahwah, NJ: Lawrence Erlbaum Pub. King, A. (2002). Structuring peer interaction to promote high-level cognitive processing. Theory into Practice, 41, 33–40. King, A., Stafﬁeri, A., & Adelgais, A. (1998). Mutual peer tutoring: Effects of structuring tutorial interaction to scaffold peer learning. Journal of Educational Psychology, 90, 134–152. Lou, Y., Abrami, P., Spence, J., Poulsen, C., Chambers, B., & d’Apollonia, S. (1996). Within-class grouping: A meta-analysis. Review of Educational Research, 66, 423–458. Mercer, N. (1996). The quality of talk in children’s collaborative activity in the classroom. Learning and Instruction, 6, 359–377. Mercer, N., Wegerif, R., & Dawes, L. (1999). Children’s talk and the development of social reasoning in the classroom. British Educational Research Journal, 25, 95–112. Meloth, M., & Deering, P. (1999). The role of the teacher in promoting cognitive processing during collaborative learning. In A. O’Donnell, & A. King (Eds.), Cognitive perspectives on peer learning (pp. 235–255). Mahwah, NJ: Lawrence Erlbaum Pub. O’Donnell, A. (1999). Structuring dyadic interaction through scripted cooperation. In A. O’Donnell, & A. King (Eds.), Cognitive perspectives on peer learning (pp. 179–196). Mahwah, NJ: Lawrence Erlbaum Pub. Otis, A., & Lennon, R. (1993). Otis–Lennon school ability test (6th ed). San Antonio, TX: Harcourt Brace. Palincsar, A. (1998). Keeping the metaphor of scaffolding fresh—A response to C. Addison Stone’s ‘‘The metaphor of scaffolding: It’s utility for the ﬁeld of learning disabilities’’. Journal of learning Disabilities, 31, 370–373. Palincsar, A., & Brown, A. (1988). Teaching and practicing thinking skills to promote comprehension in the context of group problem solving. Remedial and Special Education, 9, 53–59. Palinscar, A., & Herrenkohl, L. (2002). Designing collaborative learning contexts. Theory into Practice, 41, 26–35. Queensland School Curriculum Council. (2000). Studies of society and environment years 1–10 syllabus. Brisbane: Author. Rogoff, B., & Toma, C. (1997). Shared thinking: Community and institutional variations. Discourse Processes, 23, 471–497. Rojas-Drummond, S., & Mercer, N. (2003). Scaffolding the development of effective collaboration and learning. International Journal of Educational Research, 39, 99–111. Schultz, S., Scarloss, B., Lotan, R., Abram, P., Cohen, E., & Holthuis, N. (2000). Let’s give ‘em somethin’ to talk about: Teachers’ talk to students in open-ended group tasks. Paper presented at the AERA Annual Meeting, New Orleans, LA. Shachar, H., & Sharan, S. (1994). Talking, relating, and achieving: Effects of cooperative learning and whole-class instruction. Cognition and Instruction, 12, 313–353. Sharan, S., Shachar, H., & Levine, T. (1999). The innovative school: Organization and instruction. Westport, CT: Bergin & Garvey. Slavin, R. (1995). Cooperative learning: Theory, research, and practice (2nd ed). Boston: Allyn and Bacon. Turner, J., Midgley, C., Meyer, D., Gheen, M., Anderman, E., Kang, Y., et al. (2002). The classroom environment and students’ reports of avoidance strategies in mathematics: A multimodal study. Journal of Educational Psychology, 94, 88–106. Vygotsky, L. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University 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, & R. Schmuck (Eds.), Learning to cooperate, cooperating to learn (pp. 5–15). New York: Plenum. Webb, N. (1992). Testing a theoretical model of student interaction and learning in small groups. In R. Hertz-Lazarowitz, & N. Miller (Eds.), Interaction in cooperative groups (pp. 102–119). Cambridge, UK: Cambridge University Press. ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279278 Webb, N., & Farivar, S. (1999). Developing productive group interaction in middle school mathematics. In A. O’Donnell, & A. King (Eds.), Cognitive perspectives on peer learning (pp. 117–149). Mahwah, NJ: Lawrence Erlbaum Pub. Webb, N., Farivar, S., & Mastergeorge, A. (2002). Productive helping in cooperative groups. Theory into Practice, 41, 13–20. Webb, N., & Mastergeorge, A. (2003). Promoting effective helping behavior in peer-directed groups. International Journal of Educational Research, 39, 73–97. Wegerif, R., Mercer, N., & Dawes, L. (1999). From social interaction to individual reasoning: An empirical investigation of a possible socio-cultural model of cognitive development. Learning and Instruction, 9, 493–516. Further reading Tabachnick, B., & Fidell, L. (1996). Using multivariate statistics (2nd ed.). New York: Harper Collins. ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279 279 and the development of social reasoning in the classroom. British Educational Research Journal, 25, 95–112. Meloth, M., & Deering, P. (1999). The role of the teacher in promoting cognitive processing during collaborative learning. In A. O’Donnell, & A. King (Eds.), Cognitive perspectives on peer learning (pp. 235–255). Mahwah, NJ: Lawrence Erlbaum Pub. O’Donnell, A. (1999). Structuring dyadic interaction through scripted cooperation. In A. O’Donnell, & A. King (Eds.), Cognitive perspectives on peer learning (pp. 179–196). Mahwah, NJ: Lawrence Erlbaum Pub. Otis, A., & Lennon, R. (1993). Otis–Lennon school ability test (6th ed). San Antonio, TX: Harcourt Brace. Palincsar, A. (1998). Keeping the metaphor of scaffolding fresh—A response to C. Addison Stone’s ‘‘The metaphor of scaffolding: It’s utility for the ﬁeld of learning disabilities’’. Journal of learning Disabilities, 31, 370–373. Palincsar, A., & Brown, A. (1988). Teaching and practicing thinking skills to promote comprehension in the context of group problem solving. Remedial and Special Education, 9, 53–59. Palinscar, A., & Herrenkohl, L. (2002). Designing collaborative learning contexts. Theory into Practice, 41, 26–35. Queensland School Curriculum Council. (2000). Studies of society and environment years 1–10 syllabus. Brisbane: Author. Rogoff, B., & Toma, C. (1997). Shared thinking: Community and institutional variations. Discourse Processes, 23, 471–497. Rojas-Drummond, S., & Mercer, N. (2003). Scaffolding the development of effective collaboration and learning. International Journal of Educational Research, 39, 99–111. Schultz, S., Scarloss, B., Lotan, R., Abram, P., Cohen, E., & Holthuis, N. (2000). Let’s give ‘em somethin’ to talk about: Teachers’ talk to students in open-ended group tasks. Paper presented at the AERA Annual Meeting, New Orleans, LA. Shachar, H., & Sharan, S. (1994). Talking, relating, and achieving: Effects of cooperative learning and whole-class instruction. Cognition and Instruction, 12, 313–353. Sharan, S., Shachar, H., & Levine, T. (1999). The innovative school: Organization and instruction. Westport, CT: Bergin & Garvey. Slavin, R. (1995). Cooperative learning: Theory, research, and practice (2nd ed). Boston: Allyn and Bacon. Turner, J., Midgley, C., Meyer, D., Gheen, M., Anderman, E., Kang, Y., et al. (2002). The classroom environment and students’ reports of avoidance strategies in mathematics: A multimodal study. Journal of Educational Psychology, 94, 88–106. Vygotsky, L. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University 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, & R. Schmuck (Eds.), Learning to cooperate, cooperating to learn (pp. 5–15). New York: Plenum. Webb, N. (1992). Testing a theoretical model of student interaction and learning in small groups. In R. Hertz-Lazarowitz, & N. Miller (Eds.), Interaction in cooperative groups (pp. 102–119). Cambridge, UK: Cambridge University Press. ARTICLE IN PRESS R.M. Gillies / Int. J. Educ. Res. 41 (2004) 257–279278 Webb, N., & Farivar, S. (1999). Developing productive group interaction in middle school mathematics. In A. O’Donnell, & A. King (Eds.), Cognitive perspectives on peer learning (pp. 117–149). Mahwah, NJ: Lawrence Erlbaum Pub. Webb, N., Farivar, S., & Mastergeorge, A. (2002). Productive helping in cooperative groups. Theory into Practice, 41, 13–20. Webb, N., & Mastergeorge, A. (2003). Promoting effective helping behavior in peer-directed groups. International Journal of Educational Research, 39, 73–97. Wegerif, R., Mercer, N., & Dawes, L. (1999). From social interaction to individual reasoning: An empirical investigation of a possible socio-cultural model of cognitive development. Learning and Instruction, 9, 493–516. Further reading Tabachnick, B., & Fidell, L. (1996). Using mu