Theoretical triangulation as an approach for revealing the complexity of a classroom discussion

Stimulating collaborative reasoning in whole-class discussions

In this study we apply theoretical triangulation on a whole-class discussion in history. In their review on research in history education, VanSledright and Limón (2006) point out that in daily history classrooms it is often the teacher who does most of the talking and teacher talk, such as lecturing and story-telling, often dominates. Also in whole-class discussions, which are quite commonly used in history education, it is often the teacher who dominates the conversation (at least in history classrooms in our country, the Netherlands). Whole-class discussions can be used in the instruction phase when constructing an historical explanation or narrative (cf. Leinhardt, 1993; Halldén, 1994), but also during the debriefing phase, in which earlier (group) work is discussed. Our focus here is on the latter since whole-class discussions in the debriefing phase are especially suited for collaborative knowledge construction. When students have prepared their work individually or in groups, it enables them to participate in and make substantive contributions to the discussion. The whole-class discussion can support processes of consensus building, of making connections between the individual and the community and of transforming student findings into cultural norms (Enyedy, 2003) and may contribute to attaining a higher level of reasoning (Hogan et al., 2000; van Boxtel, 2002).

An important characteristic of whole-class discussions in which knowledge is coconstructed is that students are actively involved in the disciplinary discourse and make progression in it (Engle & Conant, 2002; emphasis added). These three notions underlie the theoretical approaches we adopted in our analysis of classroom discourse using theoretical triangulation: interactivity of the discourse, conceptual level raising and historical reasoning. Each of these approaches highlights different aspects of the discourse: the role of the teacher, of the students and of the domain, respectively. Below we elaborate on each of these perspectives.

Enhancing a highly interactive discourse

The main role of the teacher in whole-class discussions that reflect collaborative reasoning is to elicit and sustain an ongoing dialogue. Kovalainen et al. (2002) identified four modes of teacher participation during whole-class discussions in a community-of-learners classroom that support classroom interaction: (1) an evocative mode (e.g. asking stimulating questions, evoking initiation and confirmation), (2) a facilitative mode (e.g. re-voicing questions and interpretations, drawing together perspectives and initiations), (3) a collective mode (e.g. promoting collective responsibility, orchestrating discourse turns) and (4) an appreciative mode (e.g. valuing the contributions of the participants). Chin (2006) pointed out that important factors in eliciting knowledge construction are the kind of questions the teacher asks and the feedback that is provided. The purpose of the question is to elicit students' thinking, to make it explicit and open for further discussion. Feedback should function as a starting point for a new or continuous line of reasoning instead of evaluating the response against scientific or cultural norms. Scott et al. (2006) showed that in doing so a new interaction pattern (compared to the IRE- pattern) emerges, which takes the form of I-R-P-R-P-R, where P stands for prompt. The function of the prompt is then to have students explicate and elaborate their ideas. In this respect, Kovalainen and Kumpulainen (2009) make a distinction between eight types of interaction sequences, depending on who initiated (teacher or student) and how many participate (student only, teacher only, bilateral or multilateral).

Eliciting the kind of student discourse described above is not an easy task and it sets high demands on the knowledge of the teacher with regard to reasoning in the specific discipline and the difficulties students face while reasoning. The teacher should know how knowledge in the discipline is organized, recognize misconceptions and recognize opportunities for learning (cf. Cazden, 2001). Teachers encounter various problems and dilemmas, such as how to involve as many students as possible, what kind of questions promote further learning, allocation of turns and how to respond to wrong answers (cf. Cazden, 2001; Alexander, 2008). From this, it follows that a teacher has to make many decisions on how to respond to students' contributions and that different responses may have a different effect on eliciting a continuous line of reasoning. In reaction to IRE sequences, Schoenfeld (2002) identified a classroom routine that has led to a practical scheme with which interactive classroom discussions can be analyzed and different classroom discussions can be compared. The routine has as its function the elicitation and elaboration of student ideas and makes explicit the choices a teacher makes during a whole-class discussion. Examples of such choices are whether issues need to be pursued, whether clarification, expansion or reframing is necessary or whether more discussion is needed.

Conceptual level raising

When focusing on the discourse, it is important to elaborate on what kind of interactions contribute to progression and learning. Insights gained from research on student-student interaction may be helpful here. Van Boxtel (2004), for example, points out that interaction processes that contribute to learning are characterized by elaborations, questioning and solving cognitive conflicts. Dekker and Elshout-Mohr (1998, 2004) developed a process model in which they discerned three kinds of activities that may contribute to conceptual level raising in individuals during collaborative learning, namely (1) key activities, (2) regulating activities and (3) mental or cognitive activities (see Appendix C). Key activities for level raising are: to tell or show one's own work, to explain, to justify and to reconstruct one's work. Telling or showing one's own work makes one aware of one's own work (mental activity), explaining makes one think about and reflect on it and justifying makes one check it. If checking reveals sincere weaknesses, one can start to criticize one's own work and eventually reconstruct it. Reconstruction may lead to level raising. The authors point out that it may be hard to fulfil key activities for oneself and that such activities can come more naturally if there are other persons to regulate them. They discern three regulating activities: (1) asking to show one's work (2) to explain it and (3) to criticize it. Although the process model was originally developed for group work, a teacher (instead of another student) can also fulfil this regulating role.

Historical reasoning

The last perspective included in our analysis is related to the disciplinary discourse and focuses on historical reasoning. When analysing interaction in the classroom it is important to include the specific features of the domain (van Boxtel, 2004; van Drie & van Boxtel, 2008). What kind of reasoning do we have in mind when talking about the domain of history? Recent views on history learning stress the importance of students learning to reason with facts and stories from the past and learning to create new stories (cf. Perfetti et al., 1995; van Drie & van Boxtel, 2008). Historical reasoning is therefore an important goal of history education. Van Drie and van Boxtel (2008) describe historical reasoning as an activity in which a person organizes information about the past in order to describe, compare and/or explain historical phenomena. Based upon literature research and analysis of students' historical reasoning, they identified six components of historical reasoning, namely asking historical questions, use of sources, contextualization, argumentation, use of substantive concepts and use of meta-concepts. Historical reasoning in whole-class settings then implies that not only the teacher, but also the students contribute to the description, comparison and explanation of historical phenomena, and in doing so make use of the several components of historical reasoning (van Drie & van Boxtel, 2011).

Aims and research question

In this paper we demonstrate how we used theoretical triangulation as an approach to analyze a whole-class discussion in history. With this study we aim to contribute to the ongoing search for methodologies that may help us understand the complexity of classroom interactions. In our analysis we incorporate three different theoretical perspectives based on our view that in whole-class discussions students should be actively involved in the disciplinary discourse and make progression in it. These perspectives are: interactivity of the discourse, conceptual level raising and historical reasoning, thus including the teacher and the students, as well as the domain at hand. By using theoretical triangulation we aim to get a more complete view of the quality of the whole-class discussion. The research question that guides this study is: How can the quality of a whole-class discussion in history be described when using theoretical triangulation?

Perspective 1: interactive routines

To gain insight into the decisions the teacher takes in orchestrating this wholeclass discussion, we used Schoenfeld's (2002) flowchart of highly interactive routines. We discerned seven passes, which we describe below.

First pass: lines 1–7a

This first pass is shown in Figure 2. The teacher gave context and background for the topic before the students started to work in small groups. Now it is time to start the discussion about the results. In line 1 the teacher asks Marian for the result on Pip (step [A2]). In lines 2 and 4 Marian tells where her small group has placed Pip in the diagram. Her answer is on target ([D1] = no), but needs explanation ([D3] = yes). In line 5 the teacher asks for an explanation (step [A5]). In line 6 Marian gives an explanation, which needs no further explanation according to the teacher ([D3] = no). Expansion or reframing is not at stake ([D4] = no), but more discussion on the subject is needed ([D5] = yes).

Second pass: lines 7b–9a

In line 7b the teacher invites students to come up with different answers (step [A2], see Figure 3). In line 8 Cees puts forward a different answer and adds an explanation for the difference. The answer is on target ([D1] = no) and the teacher decides that no further explanation is needed ([D3] = no). But she does want more discussion on the subject ([D5] = yes).

Third pass: lines 9b–11a

In line 9b the teacher invites students to react to Cees' answer (step [A2], see Figure 4). In line 10 Sarah reacts to the place of Pip in relation to the dimension of personal interest. The teacher wants to discuss the place in relation to the other dimension, so the response raises another issue ([D1] = yes), but the teacher doesn't want to pursue this now ([D2] = no) and in line 11a she seeks closure by bringing the discussion back to the other dimension of collaboration and resistance (step [A3]). She wants more discussion about that choice ([D5] = yes).

Sixth pass: lines 15b–17

In line 15 the teacher asks Wout for his reaction (step [A2], see Figure 5). In line 16 Wout gives his reaction, which is not very clear. But it doesn't raise other issues ([D1] = no). And the teacher decides not to ask for clarification ([D3] = no). Instead she decides that a reframing of the question she wants discussed would be useful ([D4] = yes). So she reframes it in line 17 (step [A6] and [D5] = yes), which is, at the same time, the start of the final pass (step [A2]).

Seventh pass: lines 18–23

In lines 18–20 Erin, Monica and Anke discuss the question posed by the teacher. Their arguments do not raise other issues ([D1] = no, see Figure 6). But some clarification is needed ([D3] = yes), so in line 21 the teacher stimulates further reasoning (step [A5]). In line 22 Anke gives a clear explanation ([D3] = no). The teacher uses that explanation to expand on the issue ([D4] = yes). In line 23 she highlights particular aspects of the discussion, on which there is also a debate between historians (step [A6]). She decides that the question has been sufficiently discussed ([D5]) and moves on to the next subject (step [A7]).

This analysis makes clear that the teacher keeps the discussion firmly focused on the subject (several straightforward passes). She gives different students the opportunity to present their ideas. As they have worked in small groups, these ideas are clearly formulated. She only has to ask for a clarification once (first pass). When there is a ‘wrong’ answer the teacher chooses not to evaluate it herself, so no IRE. Instead she brings the answer into the discussion (second pass). She keeps the discussion on the topic. When, according to the teacher, the question related to the ‘wrong’ answer needs to be clarified and discussed, she interrupts and reframes the question (sixth pass). Then the discussion is freer and more between the students. The teacher pushes a bit towards a conclusion and gives a small concluding lecture at the end (seventh pass).

Perspective 2: conceptual level raising

The second perspective focuses on conceptual level raising. In the second column in Figure 7 the line-by-line analysis with the process model is presented.

The teacher asks several students to show and explain their work and these students do so. She deliberately gives the opportunity to show different work, which is stimulating for discussion and the process of level raising. The work shown in relation to the personal-interest axis is more or less the same. The differences are in relation to the resistance axis and the concept of collaboration. When Cees shows work that the teacher does not agree with, she does not criticize the work, but invites other students to react. She seems aware that her own evaluation would stop the discussion. That is a difference with the interaction as described in the process model. The role of criticism by fellow students is crucial for justifying. However, when a teacher gives criticism, students tend to accept this criticism without justifying because of her authority. A way out is for the teacher to organize criticism between students, and that is what she does when inviting other students to give their comments. In the discussion some students do criticize the work of others, but the key activity of justifying hardly occurs. That has to do with the fact that there is no clear norm on which to check the work. The students cannot decide whether their work is good or not. So level raising stops here, although the students feel the criticism of the teacher and start reconstructing their work a little. But it is not after trial and error in justifying, which is necessary for conceptual level raising. Eventually the teacher brings in a norm, which makes justifying possible, and she reconstructs the ‘wrong’ work of Cees and his small group herself. But it is doubtful if this will lead to level raising in the students.

Perspective 3: historical reasoning

The third perspective relates to the domain under discussion and the historical reasoning that takes place. Considering the goals of this particular task, the components use of substantive concepts (in particular the concepts resistance, collaboration, personal interest, and common good) and argumentation are most central. Although in history students often back their claims with evidence from historical sources, this is not the case here, since there are no sources included in this task. The only ‘evidence’ the students can use are the short descriptions of the people. Underlying this assignment is the role of empathy. Students have to place themselves in situations like that of Pip, and try to make sense of the choices made and evaluate these choices in terms of concepts, such as collaboration and resistance. The line-to-line analysis on historical reasoning is presented in Figure 7, third column.

The analysis shows that the following components of historical reasoning are present in this fragment: asking historical questions; argumentation; and use of substantive concepts. Not present in this fragment are the components: use of historical sources, which was expected since no sources are included in this task; contextualizing, by situating the acts of this person in a broader historical context or making direct references to dates, events or developments characteristic of that historical period; and the explicit use of meta-concepts (e.g. cause, change, argument pro, argument contra, fact). This fragment shows a lot of argumentation—students present their different points of view, provide arguments for them (by themselves, or in response to the teacher) and discuss them. The process of argumentation and the use of concepts seem closely intertwined. In their reasoning, the students make explicit use of the central substantive concepts (with the exception of ‘common good’, which all seem to agree does not apply to Pip) and a new concept is brought in (passive resistance, line 8). However, these concepts are not explicitly explained or defined here. The arguments used are often loosely based on the short description included in the task; the students do not seem to make use of their more general knowledge on this topic. The particular case of Pip is not set in the broader context of the Second World War, or related to more general issues such as why people joined the Dutch National Socialistic Party (NSB) (cf. component of contextualization). This fragment also shows the importance of the asking historical questions component, which can function as an ‘engine’ of historical reasoning (Logtenberg, van Boxtel, & van Hout-Wolters, 2011). In this example it is the teacher who raises the questions: more general questions (e.g. lines 5 and 7b), by which she elicits student reasoning, and more historical questions. In line 17 she asks whether cancelling one's membership of the National Socialistic Party is an act of resistance. By asking this question the teacher focuses the attention on the concept of resistance and brings the discussion to a more general level. Consequently, this question elicits extended student reasoning.

Combining the three perspectives: two examples

From the different analyses it becomes clear that the teacher manages to lead a whole-class discussion, giving opportunities to different students to contribute and to show and explain their work, which reveals argumentation intertwined with historical concepts. The three different perspectives used in the analyses complement each other and provide us with deeper insights into the complexity of the discussion. How these analyses complement each other, and thus what the value is of theoretical triangulation, can be illustrated by two examples.

One striking moment is when one of the students gives an unexpected answer (line 8). The teacher does not evaluate the answer herself but brings it into discussion. The analysis from the perspective of the teacher shows that through this action she realizes a continuous discourse. On the other hand from the process model, criticizing would be helpful for justifying, but only if it comes from a fellow student, not from the teacher. So her reaction is also in line with the process model, considering the fact that she is not a fellow student, but an authority. From the history perspective her reaction is also useful as it would give students an opportunity to bring in counter-arguments and use meta-concepts. But, as the analysis with the process model shows, justifying and reconstruction do not appear in the discussion and the meta-concepts that would help for justifying do not emerge. The discussion stays on a more superficial level.

A second example of the complementary aspect of the different analyses is when the teacher raises in line 17 the question of whether cancelling one's membership of the NSB would be an act of resistance. From the perspective of the teacher this is called reframing (Schoenfeld, 2002). The historical reasoning analysis provides insight in this reframing from a domain-specific perspective. The teacher reformulates the question on a more general level, by which we mean not focusing on the specific situation of Pip. In doing so, she directs the attention to whether the concept resistance is applicable to cancelling one's membership of the NSB. The reformulated question elicits a discussion in which various students participate. However, with respect to the historical reasoning component using substantive concepts, the teacher could also have elaborated and deepened the discussion by explicitly asking the meaning of the concept resistance. Now, the meaning of this concept remains more or less implicit. In addition, the analysis from the perspective of the students reveals that a clear norm for justification is still lacking, so conceptual level raising does not take place. Thus, although the teacher manages to sustain an ongoing dialogue in which several students participate, the quality of the students' responses stays on a (more or less) superficial level.

These examples make clear that combining these three different perspectives provides us with a deeper insight in this piece of classroom discourse. It shows that some actions of the teacher may be useful from one perspective, but less from another. It thus reveals the complexity of eliciting collaborative knowledge construction in whole-class discussions.