You are here

Good talk in the classroom: a collaborative classroom research project

Alan Cox, Brian Finch, and Winifred Jackson

The Good Talk project investigated how to develop children’s skills more effectively in the Interpersonal Speaking and Listening Achievement Objectives in English, and in some of the Essential Skill areas, through strategies for improving group discussion. The results have useful implications for teaching. They show that children can reflect seriously on their own learning processes, through “talk about talk”. They can also engage enthusiastically with curriculum content ideas, through “talk about topic”. However, it seems to be difficult for them to do both at once.

Journal issue: 

Good talk in the classroom
a collaborative classroom research project

Alan Cox, Brian Finch and Winifred Jackson

Massey University College of Education

The Good Talk project aimed to investigate how to develop children’s skills more effectively in the Interpersonal Speaking and Listening Achievement Objectives in English in the New Zealand Curriculum (Ministry of Education, 1994), and in some of the Essential Skill areas detailed in The New Zealand Curriculum Framework (Ministry of Education, 1993). Early discussions focused on research such as the National Oracy Project in Britain (Norman, 1992), with particular focus on the value of children’s reflecting on their own learning. “Reflecting, including the reflection that is enabled by talk outside the event, seems to be an essential pre-requisite for critical thinking and the modification of what we believe” (Barnes, 1992, p.127).

The project involved two Year 5-6 classes. At early planning meetings, the team of six (three lecturers, the school principal, and two teachers) established the focus and methodology for the Collaborative Classroom project. The project followed action research principles of collaboration, where teachers conduct investigations focusing on their own practice (Poskitt, 1995), by including other individuals who jointly share the research (McTaggart, 1991).

The Good Talk team planned for regular opportunities for group discussion and whole-class reflection to be included in the two classes’ normal programmes. These would be the focus of observations by the lecturers over a period of twelve weeks. Regular reflective discussion of the observation by the college-school team determined the subsequent direction of the project. It was convenient for the project that the two teachers worked closely together planning their programmes.

The team worked from the general premise that an ability to reflect on processes of learning is an important step in improving the learner’s own skills and effectiveness. Studies in how metacognitive awareness positively affects how pupils learn (Nuthall and Alton-Lee, 1994) indicate that this awareness develops “in interaction with other people”. The research team felt that by helping children reflect on their group discussions, the children were more likely to develop effective and democratic speaking and listening skills.

Specifically, the team wanted the children to be aware of the processes and patterns of interaction within their group discussions. The information letter stated: “The children’s own thoughts about the project will be an important part of the evaluating process, and they will be fully informed of progress throughout the project… the children will have the opportunity to add their own observations to the data being gathered. They will be important partners in the research.” The children had the nature of the research topic, and their role in it, explained to them at the time their permission was sought (Hammack, 1997).

The form that the project took, in response to initial and ongoing observations, was first to implement changes to group composition and group sizes. Then the children re-established their own rules for good talk in group discussions. Strategies were implemented to make children more aware of their own and others’ participation in group-work situations. Opportunities were provided for group and whole-class discussion and evaluation of their work in groups. A questionnaire was completed early in the project and again at project end to provide data on attitudes to group work.

Six strategies for improving group discussion were considered by the college-school team, but because of time restraints, two of these were not trialled.1 The children’s involvement was thoughtful and serious as they completed the “before and after” written questionnaire and gave their verbal reflections on the four trial strategies selected for group discussion: Making the Rules, Play the Card, The Pupil Monitor, and Making Connections. The children seemed interested in the central question of what factors make for good talk.

The team decided to involve the children in the “Making the Rules” exercise as a way of drawing on their own prior experience of working in groups and, at the same time, increasing their awareness of the focus on good talk. Groups drew up their own rules for holding group discussion, and then the class worked out a list acceptable to them all. This exercise, together with the children’s responses to the questionnaire, established a valuable threshold for the project and was central to later interesting reflection on group talk by both the team and the children. Their early perceived “givens” for successful group talk were:

•&&Don’t put others’ ideas down

•&&Build on others’ ideas

•&&Let everyone have a turn

•&&Speak clearly/use “inside” voices

(i.e. not playground voices)/use eye contact

•&&Stay on the subject

•&&Listen to others’ ideas

•&&Give constructive criticism.

A notable feature was the later realisation that these rules were often ignored in the subsequent cut and thrust of heated discussion, when the topic was one close to their hearts. Later observed sessions, e.g. when the groups worked with non-participating Pupil Monitors, brought this discrepancy between pupils’ perceptions and the reality into focus. An amusing anecdote illustrates this. A serious boy was observed meticulously fulfilling his role as monitor. He took care to note every utterance while the group of four, three girls and a boy, enthusiastically debated the question “What is the best fast food and where can you get it?”. The talk was dominated by a strong-minded girl, keen to say her piece, and not inclined to listen to the others. When the monitor reported his results, revealing that she had hijacked the discussion, she was irate: he had “faked the results and better change them”! The ideal of democratic give and take, sincerely established in “Making the Rules”, had clearly taken second place to assertive individualism.

With hindsight, the observers felt that having a pupil monitor had little immediate effect on the dynamics of the group, but it provided a valuable opportunity for the monitor to “sit back” and reflect on the nature of “good talk” without having to take sides in the competitive hurly-burly of the discussion. In the long run, these insights of the monitors (all would be given the responsibility to be a monitor in subsequent group discussions) could eventually bear fruit in more effective group talk. Certainly, the results of the second questionnaire indicated this. Those who had expressed positive attitudes towards involvement in group discussions retained those attitudes, and others expressed significant shifts to more positive attitudes, e.g. in terms of managing to have their contribution heard.

Implications for teaching

Observations were focused on the verbal interaction within the group, but the observers also noted any other factors which seemed to affect interactions. Recurrent patterns identified a number of implications for teachers in the areas of gender, group size and group dynamics. Observations of the use of the introduced strategies and differences in discussion topics also provided material from which implications could be drawn.



Gender interaction was strongly influenced by the seating groups in use at the time. When the study began the students were seated in same-gender groups and there was very little oral communication between boys and girls. When seating was changed to mixed gender groups, comfortable oral interaction was observed between boys and girls in those groupings.

Implication: Teacher organisation is important for creating opportunities for natural social interaction between boys and girls.

Point of view

Observation of problem-solving discussions showed that the girls would provide their own viewpoint and often that of someone in their family (eg. Mum). The boys’ discussions tended to rest solely on how the situation would affect them personally.

Implication: Awareness of other viewpoints could be an area to work on, and one which would also support group work skills.

(c) Competitiveness

This was observed among all-boy groups, e.g. competing to finish first, or to be the leader. This kind of competitiveness had a negative effect on what those individuals gained from the task.

Implication: Emphasising the learning objectives of the larger task may help to overcome any tendency to rush or compete to finish smaller parts of it.


(a) Size

A range of group sizes, from pairs to groups of six, was observed. Paired discussions were particularly effective in giving the more reticent children the opportunity to be significant participants. However, any discussions were negatively affected by one of the pair dominating the task agenda or the resource materials, thereby blocking the partner out of the proceedings. Where neither pupil was able to bring prior skills or understanding to the discussion, pair work was unproductive.

Groups of three were mostly efficient in structuring the decision-making process, and kept well-focused on the topic. The 2:1 gender mix was not seen as a significant factor. The talk was equitably shared. But groups of four and five were seen to be the most effective, particularly when the children sat in a square or circle with faces no more than about a metre apart. Gender factors were also significant in these groups. Participation levels tended to be equitable in groups with 3:2 or 2:2 ratios, but generally a 3:1 or 4:1 ratio meant that the single boy or the single girl contributed very little, and even became physically distanced from the group. The most even oral participation occurred in the 2:2 groups, where neither gender was outnumbered.

Implication: Choice of group size is a real factor in facilitating good talk and composition. Gender balance was a factor with these Y5-Y6 students.

(b) Desk geography

The physical position of the desks was a factor; for example, in a group of five it was easy for one student to drift out of the discussion. In one observation, a girl who was “monitor” (a neutral role) showed she was aware that one of the participating members was located out of the discussion, and changed places with her.

Groups of six lacked equitable participation. They tended to be restrained by the furniture, which impeded the formation of a circular group. Individuals were as much as 3 metres apart, and one or two of the group easily became non-participants; or else there was a tendency for the group to become two groups of three, sometimes split on gender lines.

Implication: Desk placement needs to be considered to ensure that proximity and seating positions allow for easy audibility and eye-contact.

(c) Body language

The observations showed that students’ body language reflected the amount of involvement. For example, when heads were close and leaning in to a centre, the students were absorbed in the task. Eye contact was another indication of effective discussion.

Implication: Some children would benefit from learning to make more use of eye contact during group interactions. For example, modelling the scanning of a group would show students the practical value of seeing who has ideas to share, and also of making all members of the group feel included.

Group dynamics

(a) Roles in the group

Groups were often subverted by one or two of the stronger personalities who tended to say the most, listen the least, and take over the decision-making. This dominance was often reinforced by the stronger-willed pupils acting as “power-brokers” by assuming the controlling roles of unofficial chairman, interpreter of task instructions, recorder or reporter. It was observed that the role of “recorder” in a group could influence the dynamics of groups in different ways; for example, when the recorder participated in the talk, little recording was done; discussion then stopped while the recorder wrote.

Implication: To become confident in taking on positions of responsibility, children should be rostered to take a range of group roles, and be given strategies for summarising data in brief – for example, by using an appropriate record-sheet with pre-determined sub-headings.

(b) Compromise

One group was observed in which all four held different opinions and did not know how to progress further. Some students were observed who preferred to lead, but were not good at compromise, and lost interest when their ideas were not taken up.

Implication: Some pre-teaching and modelling of strategies for working towards consensus could be valuable.

(c) Leadership

The group dynamic was also influenced by the combinations of individuals who composed the group. Generally, the students who were very able were adaptable to roles in group work and could work with other children. However, children were never observed drawing in the group members who were not participating.

Implication: Students who remain quiet in group discussion do not get practice in the leadership role. Strategies are needed to draw all members of a group into discussion.

Introduced strategies

The team observed the students making use of the strategies discussed below, and becoming more aware of their own contribution and the dynamics of group discussion.

(a) Play the Card

(Speakers play a coloured card each time they contribute)

Strict adherence to the principle of turn-taking was seen to have a negative effect on occasions, e.g. one child became upset when his group noticed he had two more turns than the others (although he then argued that he had more things to say). Talking in strict turns engendered talk for the sake of it. However, when playing the cards without strict turn-taking being required, students’ contributions were more significant. Children were also aware of when someone was not contributing. Reflection enabled some students to be aware of how constructive comments can inject vigour back into flagging discussions.

Implication: The cards can make the share of talk visible, make students aware of their contribution and motivate all members of the group to contribute.

(b) The Pupil Monitor

(One group member is a non-participating observer)

Children managed this well. The groups observed seemed more natural in their group interactions and were less bound by turn-taking. One monitor commented perceptively to his group on how there had been a lot of interrupting of one speaker. Children were surprised to learn that their actual level of participation was not what they had thought. They had constructive ideas to improve discussion. For example, they suggested that a card should be played for a new idea, rather than for a mere utterance; that discussion time should be limited to help maintain focus; and that observer-monitors should be allocated time for feedback.

Implication: The use of a non-participating monitor needed to be allowed for in group size; groups of three suffered the same problems as pairs when one of the three was acting as monitor. This monitoring strategy increased students’ awareness of their own and others’ share of talk.

(c) Making Connections

(Always begin by linking to someone’s prior contribution)

This was introduced towards the end of the study. From observation, it did not seem to get beyond the formulaic “I agree with so and so”. The idea of not copying others’ ideas was voiced a few times, and it may be that this impeded the development of an idea put forward by someone else.

Implication: There was a need for more modelling of this strategy and more variety of approaches to draw ideas out. A video of effective and ineffective techniques would be a useful resource for teachers.

Choice of topic

Topic discussion was most productive when children had some prior knowledge from earlier lessons or their own experience. Observations of children with little prior knowledge showed them not taking part in the talk and sometimes working inefficiently on the task.

Implication: If the topic draws on general knowledge, the teacher may need to provide the background resources so that students start discussion on a more equal footing.

The challenge: two kinds of talk

As the project progressed, the team recognised that the children’s talk could be categorised as either “talk about talk” or “talk about topic”. The children used “talk about talk” when trialling the speaking and listening strategies and reflecting on them. They were, in effect, being asked to be co-researchers, focused on the same fundamental question as their teachers and adult visitors. Some lessons were totally devoted to this “talk about talk”, e.g. when the children were formulating their own “Rules for Class Discussions”. An area of challenge noted here was that the children often “forgot” many of their good talk ground-rules, and were unaware that they were breaking the very rules they had agreed to. Situations were observed where the task was particularly absorbing and children participating were seen to shout each other down and not consider others’ ideas. The monitoring strategies brought some of these anomalies to the children’s attention, even though not without dispute.

The children were also required to focus on “talk about topic” where the lesson was on such subjects as the functional uses of water, or the pros and cons of eating fast foods. The observation made here was that a discussion could be extended for no purpose other than that of individuals talking, but without adding new input or direction to the discussion. Children noted this themselves, and suggested that time limits could remedy the problem.

Often a lesson involved both types of talk, for example, when discussion of a topic was followed by the group monitors reporting back to the class on how the discussion of their groups had progressed. It was during these evaluation sessions that it became clear that there was a difficulty in giving both these types of talk their due. The “talk about talk” showed the children to be very capable of reflection; but sometimes, when it was most successful, the results of the topic discussion would be ignored.

There is an important implication for teachers here. Children can reflect seriously on their own learning processes; they can also engage enthusiastically with curriculum content ideas. However, it seems to be difficult for them to do both at once. If the focus is strongly on how we talk, the actual topic can be trivialised, treated merely as a vehicle for the group interaction in the spotlight. If the focus is purely on the ideas, with no concern for the way all of the children are engaged in generating those ideas, then important social and communication skills are undervalued.

The team expected that the project would foster growth in the following aspects of the Framework: Communication Skills, Problem-solving, and Social and Co-operative Skills. However, it became clear that the first focus would need to be in the area of Social and Cooperative Skills, as they are so relevant to the behaviours that are needed within a group for all members to be able to participate and be valued.


This project gives useful reference points for future professional practice, and research into improving children’s purposeful classroom talk. The pupils’ “talk about talk”, and the results of the surveys, both showed that the children had responded positively to the exercise in increasing their awareness of talking behaviour in groups. They had developed in metacognitive understanding of their own speaking and listening processes.

Observation tended to support the pedagogical premise that children learn better when they know how they learn. By exploring strategies for group discussion, and by talking about them, their group talk became more effective and equitable. This was revealed by the positive shifts shown in the “before and after” questionnaire data. The opportunities for reflective discussion were “real” learning contexts because they drew on the children’s own observations.

The findings on the significance of group size, composition and physical positioning, and the need to plan for inclusion, reaffirm sound pedagogical understandings. Teachers should carefully consider these four factors.

The key challenge for the classroom teacher is to arrange an equitable balance between “talk talk” and “topic talk”. A long-term approach to building up children’s awareness may be a way to achieve this balance. In planning for a whole year’s programme in speaking and listening, teachers could attempt to resolve this dilemma by having the children focus on trialling three or four group-talk strategies in Terms One and Two. Various effective and ineffective group formations could be modelled, observed and discussed. For instance, the Fishbowl method could be used, so that children could observe others role-playing different situations. Group discussions could be video-taped to permit follow-up reflection.

Term Three could then be devoted to a series of group discussions, with no focus on using strategies, so that the teacher can observe whether the children have taken the “good-talk” strategies on board. Teacher-pupil reflection could then re-focus on the process, in preparation for Term Four discussions. Ideally, by then the children would be using the strategies as useful facilitating tools in their exploration of the topic in hand, and “good talk” would have become habitual.


1.&&For details of the six strategies, Making the Rules, Play the Card, The Pupil Monitor, Making Connections, Fishbowl, and Diagram Discussion, contact Alan Cox, Department of Arts and Language Education, Massey University College of Education, Private Bag 11-222, Palmerston North.


Barnes, D. (1992). The role of talk in learning. In K.Norman (ed.), Thinking voices: The work of the National Oracy Project. London: Hodder and Stoughton.

Hammack, F. (1997). Ethical issues in teacher research. Teachers College Record, 99 (2), pp. 247-265.

McTaggart, R. (1991). Principles for participatory action research. Adult Education Quarterly, 41 (3) pp.168-187.

Ministry of Education (1993). The New Zealand Curriculum Framework. Wellington: Ministry of Education.

Ministry of Education (1994). English in the New Zealand Curriculum. Wellington: Ministry of Education.

Nuthall, G. and Alton-Lee, A. (1994). How pupils learn. set 1, Item 3.

Poskitt, J. (1995). Action research: Assisting teacher development. NZ Principal, June, pp. 12-14.

ALAN COX, BRIAN FINCH AND WINIFRED JACKSON are all lecturers in the Department of Arts and Language Education at Massey University College of Education. They share an interest in developing language skills in schools. Their three co-researchers in this participatory Action Research Project were the principal and two teachers of parallel Year 5-6 classes in a primary school. This project was funded by the Institute for Professional Development and Educational Research, at Massey University College of Education.