Interactive Whole Class Teaching

British Educational Research Journal, Vol. 30, No. 3 June 2004

Interactive whole class teaching in the National Literacy and Numercy Strategies Fay Smith*, Frank Hardman, Kate Wall and Maria Mroz University of Newcastle upon Tyne, UK (Received 4 March 2003; conditionally accepted 23 June 2003; accepted 8 July 2003) The study set out to investigate the impact of the official endorsement of ‘interactive whole class teaching’ on the interaction and discourse styles of primary teachers while teaching the National Literacy and Numeracy Strategies. In both strategies, interactive whole class teaching is seen as an ‘active teaching’ model promoting high quality dialogue and discussion between teachers and pupils. Pupils are expected to play an active part in discussion by asking questions, contributing ideas and explaining and demonstrating their thinking to the class. Using computerized systematic classroom observation, discourse analysis of transcripts and a questionnaire, the project looked specifically at the discourse strategies currently used by a national sample of primary teachers when teaching the literacy and numeracy strategies and their perceptions of current practices. The findings suggest that traditional patterns of whole class interaction have not been dramatically transformed by the strategies. The implications of the findings for classroom pedagogy, teachers’ professional development and future research priorities are considered.

Introduction Since 1997, a major thrust of the new government has been to address standards of literacy and numeracy in English primary schools. In a bid to achieve this end, the National Literacy Strategy (NLS) was launched in 1998 (Department for Education and Employment [DfEE], 1998) and the National Numeracy Strategy (NNS) in 1999 (DfEE, 1999a). It is claimed that these policy-led initiatives have had a major impact on many aspects of primary education, including teaching styles, and there is much rhetoric about their efficacy from politicians, government agencies, the media, teachers and teachers’ representatives. A major feature of the strategies has been an emphasis on direct, ‘interactive whole class teaching’, drawing mainly on the school effectiveness and school improvement literature (e.g. Reynolds & Farrell, 1996; Reynolds 1998; Reynolds & Muijs, 1999). It is suggested that more interactive forms of whole class teaching will play a vital role in raising literacy and numeracy standards by promoting high quality *Corresponding author: School of Education, Communication and Language Sciences, University of Newcastle upon Tyne, Joseph Cowen House, St Thomas Street, Newcastle upon Tyne NE1 7RU, UK. Email: [email protected] ISSN 0141-1926 (print)/ISSN 1469-3518 (online)/04/030395-17  2004 British Educational Research Association DOI: 10.1080/01411920410001689706

396

F. Smith et al.

dialogue and discussion and raising inclusion, understanding and learning performance. In the NLS Framework, successful teaching is described as ‘discursive, characterised by high quality oral work’ and ‘interactive, encouraging, expecting and extending pupils’ contributions’ (DfEE, 1998, p. 8). Similarly, the NNS Framework states: ‘high-quality direct teaching is oral, interactive and lively … in which pupils are expected to play an active part by answering questions, contributing points to discussion, and explaining and demonstrating their methods to the class’ (DfEE, 1999a, p. 11). In both strategies, interactive whole class teaching is not seen as a return to a traditional ‘lecturing and drill’ approach in which pupils remain passive, but as an ‘active teaching’ model encouraging a two-way process. However, critics argue there is no clear definition and little practical advice for teachers on what interactive whole class teaching is and how it should be used in the classroom. For example, Galton et al. (1999) argue that little evidence has been presented to show it differs from traditional whole class teaching as reported in earlier studies of the primary English classroom (e.g. Mortimore et al., 1988; Pollard et al., 1994; Alexander et al., 1996). Studies of classroom discourse from North America and the UK (e.g. Mehan, 1979; Edwards & Westgate, 1994) show that whole class teaching across all stages of schooling is dominated by what Tharp and Gallimore (1988) call the ‘recitation script’. In its prototypical form teacher-led recitation consists of three moves: an initiation, usually in the form of a teacher question, a response in which a student attempts to answer the question, and a follow-up move, in which the teacher provides some form of feedback (very often in the form of an evaluation) to the pupil’s response. This three-part exchange structure, as revealed by Sinclair and Coulthard (1975), is therefore particularly prevalent in directive forms of teaching and consists of a series of unrelated teacher questions that require convergent factual answers and pupil display of (presumably) known information. Recitation questioning therefore seeks predictable correct answers and only rarely are teachers’ questions used to assist pupils to more complete or elaborated ideas. Brown et al. (1998, pp. 370–371) suggest that despite the claims for whole class teaching, the evidence that it is associated with higher attainment in mathematics is ‘not unambiguous’. Brown and her colleagues point out that while some studies show correlations between whole class teaching and attainment, there is also evidence that whole class teaching can be associated with particularly poor results. They go on to suggest that the quality of teacher–pupil interaction is a much more important factor than class organization, concluding that ‘a whole class format may make better use of high quality teaching, but may equally increase the negative effect of lower quality interaction’. Given the lack of empirical evidence showing that whole class teaching in the literacy and numeracy strategies is more interactive and promotes quality dialogue and discussion, this study set out to investigate patterns of whole class interaction in the NLS and NNS. We also explored whether there are differences in the discourse strategies used by effective teachers of literacy and numeracy and whether teachers varied their discourse strategies when teaching the two subjects and across the key stages (KS).

Interactive whole class teaching

397

Method Sample A national sample of 72 primary school teachers working in a range of socio-economic settings across the regions of England was selected (35 literacy, 37 numeracy). Within each subject area (literacy and numeracy) half the teachers were selected because they were highly effective: the other half made average progress with their pupils. The effectiveness of each teacher was established using value-added data (residual measures) based on Performance Indicators in Primary Schools (PIPS) data provided by the Curriculum, Evaluation and Management (CEM) Centre at Durham University (Tymms, 1999). Structuring the sample in this way allowed for an investigation into whether effective teachers are employing a different range of discourse strategies in comparison with the ‘average’ teachers (i.e. those whose value-added scores are broadly zero). Systematic sampling (a form of probability sampling) was used to select teachers from within our larger sample of effective and average teachers (i.e. selecting every nth case). Computer assisted systematic observation Observations were carried out using a computerized observation schedule developed by the research team known as the Classroom Interaction System (Smith & Hardman, 2003). A continuous sampling method was used. The coding scheme uses ‘The Observer’ software (Noldus Information Technology, 1995) to log the number of different types of discourse moves made by teachers and pupils. This was done using a hand-held device about the size of a calculator. This computerized system enabled us to observe the lesson in real-time and was quicker than traditional paper and pencil methods because the data were instantly stored, and therefore available for immediate analysis. We obtained good measures of inter-rater and intra-rater reliability (correlations of 0.86 and 0.78 respectively): an in-depth discussion of the Classroom Interaction System can be found in Smith and Hardmah (2003). The computerized system logged (for each teaching exchange): the actor, the discourse move and who the receiver was. It therefore primarily focused on the three-part, Initiation–Response–Feedback (IRF), structure first identified by Sinclair and Coulthard (1975) and gathers data on teachers’ questions, whether questions were answered (and by whom), and the types of evaluation given in response to answers. It also recorded pupil initiations in the form of questions and statements. The system recorded whether teacher questions were open (i.e. defined in terms of the teacher’s reaction to the pupils’ answer: only if the teacher will accept more than one answer to the question would it be judged as open) or closed (i.e. calling for a single response or offering facts). Responses were coded according to whether a boy or girl answered or whether there was a choral reply. Teacher feedback to a pupil’s answer was coded according to whether it was praised, criticized, or accepted. The system also captured two alternative strategies in the feedback move: probes (where the teacher stayed with the same child to ask further questions) and uptake questions (where the teacher incorporated a pupil’s answer into a subsequent question).

398

F. Smith et al.

Transcript analysis Video recording of a sub-sample of 14 effective teachers identified by the valueadded data were collected. The sample was made up of 8 Reception/KS1 teachers (4 literacy/4 numeracy) and 6 KS2 teachers (3 literacy/3 numeracy). The video recordings were transcribed and coded using an intensive system of discourse analysis adapted from the work of Sinclair and Coulthard (1975). The teaching exchanges were quantified and turned into percentage scores to compare the patterning of the teacher/pupil interactions across all 14 lessons. By focusing on the three-part, IRF structure, the findings of the discourse analysis could be triangulated with the computerized observation data. Teachers’ questions were also analysed according to whether they were open or closed and for the use of probe and uptake questions by the teacher. The average length of pupil utterances was also calculated. Comparisons across subjects and key stages were made to see if there was any variation in the type of questions asked by teachers and the length of pupil utterances.

Teacher questionnaire A self-completion questionnaire was designed to explore teachers’ understanding of the concept of ‘interactive whole class’ teaching and their perceptions of the range of discourse strategies they currently use when teaching the literacy or numeracy strategies. The questionnaire also explored their views on the quality of the training in whole class teaching they had received. Although mainly quantitative in design, the questionnaire included questions allowing for more open responses. The results of this questionnaire are reported elsewhere (Hardman et al., 2003a), but are referred to later.

Findings Computer-assisted observations Sample characteristics. Seventy-two lessons from across England were observed: 35 literacy and 37 numeracy. Roughly one-third of the lessons fell into each of Reception, KS1 and KS2. Some 60% of the lessons were taught by highly effective teachers (as classified by the value-added data). Those teachers with a value-added score above 2 were classified as being highly effective; 40% of the teachers were average—their value-added scores fell between ⫺ 0.5 and ⫹ 0.5. Table 1 shows the school and class (data in italics) statistics for the sample. The average roll call for the schools in our sample was 315, with 20% of the pupils being eligible for free school meals. On average a fifth of the pupils were on the special needs register and 10% had English as an additional language. The average class size we observed was 26—we also noted the number of males and females present in each class.

Interactive whole class teaching

399

Table 1. School and class statistics

School roll call (n) Percentage eligible for free school meals Percentage on special needs register Percentage with English as an additional language Males (n) Females (n) Class size (n)

Minimum

Maximum

Mean

SD

99 2 2 0

501 62 49 99

315.71 20.15 21.01 9.79

119 17.32 9.73 26.92

7 6 16

20 20 33

13.19 12.42 25.58

3.02 3.23 4.19

Lesson duration. On average the lessons lasted 53 minutes, and the whole class section lasted 32 minutes (60% of the lesson). Some lessons were entirely made up of whole class teaching. Our observations focused upon the whole class section only. The percentage of each lesson consisting of whole class teaching varied from teacher to teacher, as shown in Figure 1. There was no significant difference in the amount of whole class teaching taking

Figure 1. Variation in amount of whole class teaching

400

F. Smith et al.

Figure 2. Number of lessons in which the different discourse moves were observed

place in literacy and numeracy lessons. The KS2 lessons were usually about 4 minutes longer than the KS1 lessons, which were in turn 5 minutes longer than Reception lessons. Discourse strategies of the whole sample. Figure 2 shows the number of lessons in which we observed certain types of discourse. The maximum height for each bar is 72 (the number of lessons observed). Some 43% of the teachers did not use an uptake question at any time during the whole class section of the lesson. Spontaneous contributions from pupils occurred in 59 of the lessons (82%). Interruptions to the lesson occurred 42% of the time (normally other staff collecting details about dinner money and packed lunches). Also of interest is that 15% of the teachers did not ask any open questions during the whole class section (all 72 teachers asked closed questions). Figure 3 shows the rate (number per hour) for each discourse move. Rate is calculated as frequency per hour to make this data comparable to similar studies: therefore, if a teacher used 23 closed questions over a 30-minute long whole class section, this would be reported as 46 closed questions per hour. Clearly, the most frequent include closed questions (69 per hr), evaluation (65 per hr), explaining (50 per hr) and direction (39 per hr). Throughout the observations our focus was upon the teacher, but we also analysed responses and initiations from pupils during the whole class sections of the lessons. The pupils in our sample did not use any of the discourse moves in Figure

Interactive whole class teaching

401

Figure 3. Rate of teacher discourse moves

3. When pupils spoke, the most dominant discourse was to answer a question—the moves are listed as follows: • • • •

Answering a question (118 moves per hour); Choral response (13 moves per hour); Presentation (13 moves per hour); and Spontaneous contribution (9 moves per hour).

Rather than looking at rate per hour (which takes no account of the length of a discourse move), it is also possible to report the mean duration for each discourse move (average length in seconds) and the percentage duration for each discourse move (each discourse move’s total contribution to the entire whole class section, e.g. if explaining took up 5 minutes of a 20 minute whole class section the percentage duration would be 25%). Mean durations (in seconds) and percentage durations for each discourse move are shown in Table 2. The pupil discourse moves are italicized in the table. Explaining and directing were quite frequent discourse moves—as mentioned earlier. The data in Table 2 also show that these moves lasted the longest.

402

F. Smith et al. Table 2. Mean duration and percentage duration for each discourse move Discourse move Direct Explain Open question Closed question Repeat question Uptake question Probe (question) Evaluate Refocus General talk Pupil answers Choral response Spontaneous contribution Presents Interruption Total

Mean duration (secs)

Percentage duration

17.1 20.3 7.6 4.9 7.6 5.2 4.1 5.4 8.4 7.3 5.0 14.6 7.0 14.2 19.9

15.3 27.9 2.5 9.0 3.0 0.7 2.0 8.8 3.7 1.4 15.1 4.3 1.7 4.2 0.5 100

Explaining was of the longest duration (20 secs), followed by direction (17 secs). Closed questions and evaluation were frequent (as found earlier), but these moves lasted about the same amount of time as the other moves. The average length of a pupil answer was 5 seconds. Choral responses took much longer—14.6 seconds. In Table 2, the total contribution of each discourse move adds up to 100% (representing the entire whole class section of the lesson). Figure 4 shows this information more clearly. Here we can see that explaining (which was both frequent and long) took up 28% of the whole class section. Some 15% of the whole class section consisted of direction from the teacher, and another 15% of individual pupil answers. After that, closed questions and evaluation contributed the most to the whole class section. By adding up the teacher discourse moves (top 10 in the table), it is clear that the teacher dominated the whole class section for 74% of time. The 24% pupil contribution was mainly made up of answering questions individually or in choral response. Interruptions to lessons accounted for the small remaining percentage (0.5%). Literacy and numeracy lessons compared. There are clear differences in the rate (number per hour) for each discourse move between subject areas. An individual t-test found that significantly more direction took place in numeracy lessons compared to literacy lessons (t ⫽ 5.05, p ⬍ .001). On average there were 24 more direction moves in a numeracy lesson. Closed questions and choral responses were also more common in numeracy than in literacy lessons (p ⬍ .01). Uptake questions were rare in both subject areas, but more likely to happen in literacy lessons (t ⫽ 3.22, p ⬍ .01).

Interactive whole class teaching

403

Figure 4. Contribution of each discourse move to the whole class section

Although no difference was found in the number of explanations per hour in literacy compared to numeracy, explanations were found to last 5.3 seconds longer in literacy compared to numeracy lessons (t ⫽ 3.59, p ⬍ .01). This was the only significant difference found when analysing mean duration. The percentage duration of each discourse move (split by subject) is shown in Figure 5. Direction (p ⬍ .001), closed questions and choral responses (p ⬍ .01) all contributed more (in terms of percentage of the whole class section) in numeracy lessons compared to literacy lessons. Direction contributed to 12% of literacy lessons compared to 18% in numeracy lessons. Explaining, uptake questions (p ⬍ .001) and open questions (p ⬍ .05) took up more time in literacy lessons compared to numeracy. Some 33% of a literacy lesson consisted of explaining compared to 23% of a numeracy lesson. Key stage comparison. This analysis then focused upon differences in discourse across the key stages (Reception, KS1 and KS2). Figure 6 shows the differences in rate (n per hr) across the key stages. Choral responses were more frequent in Reception (24 per hr) compared to KS1 (13 per hr) and KS2 (only 3 per hr). A one-way ANOVA found this to be significant at p ⬍ .001. Uptake questions were rare in Reception year and closed questions were more common in Reception/KS1 compared to KS2

404

F. Smith et al.

Figure 5. Contribution of each discourse move to the whole class section—literacy/numeracy comparison

(p ⬍ .01). Other differences were significant at p ⬍ .05—direction, probes, refocus and general talk. Only one significant difference was found for mean duration—pupils presented for longer in KS2 than the younger year groups (F ⫽ 7.28, p ⬍ .01). A pupil presentation would normally last 21 seconds in KS2, but only 10–11 seconds in Reception and KS1. The percentage durations for each discourse move (across key stage) are shown in Figure 7. Choral responses contributed less in KS2 than the younger year groups (p ⬍ .001). General talk was more prominent in Reception year (p ⬍ .05) and refocusing also took up more time in the younger year groups (p ⬍ .05). Effective and average teachers compared. Discourse strategies were compared between the two sets of teachers in our sample—those classified as highly effective and those classified as average. Figure 8 shows the differences in rate (n per hr) between these two sets of teachers. Only one discourse move (general talk) was significantly different between the two groups of teachers. It was more frequent among the highly effective teachers (p ⬍ .05), suggesting they create opportunities for more informal talk. Further inspection of Figure 8 reveals that there seem to be more discourse moves, in general, among the highly effective teachers. To examine this further, a new variable was created which essentially was the sum total of all of the discourse moves made by a teacher and the pupils (but excluding interruption). This sum total was then divided by the duration of the whole class section of the lesson to obtain

Interactive whole class teaching

405

Figure 6. Rate of discourse moves compared across key stages

rate per hour. This variable can be seen as an indicator of pace in the lesson. A independent t-test found that highly effective teachers’ lessons consist of significantly more discourse moves than average teachers (t ⫽ 2.08, p ⬍ .05). Highly effective lessons consist of, on average, 55 more discourse moves per hour (469 moves for highly effective teachers and 414 moves for average teachers). The average length of each discourse move did not differ significantly between the two groups of teachers. Also, the percentage durations for each discourse move did not differ between the two groups of teachers.

Results from the analysis of lesson transcripts A sub-sample of 14 lessons was analysed using discourse analysis (as described earlier). The discourse analysis framework provided a clear and systematic basis for analysing the classroom discourse in all 14 lessons and for triangulating the results with the computerized systematic observation data. The qualification and subsequent patterning of the teaching exchanges provided a useful means of further exploring the discourse strategies used by teachers: when teaching literacy and numeracy lessons and across the key stages. In addition to analysing the teaching exchanges, we also explored and quantified the teachers’ use of open and closed questions, uptakes and probes in response to pupils’ answers, and the length of pupil utterances.

406

F. Smith et al.

Figure 7. Contribution of each discourse move to the whole class section—key stage comparison

Figure 9 shows the patterning of the teaching exchanges based on the percentage scores for the seven literacy and seven numeracy lessons. The analysis shows there was little overall variation in the patterning of the teacher

Figure 8. Rate of discourse moves compared across teacher effectiveness

Interactive whole class teaching

407

Figure 9. Overall analysis of teaching exchanges Key: T. In. ⫽ teacher inform, T. Dt. ⫽ teacher direct, T. El. ⫽ teacher elicit, P. El. ⫽ pupil elict, P. Inf. ⫽ pupil inform, Re-in. ⫽ re-initiate, List ⫽ listing, Rein. ⫽ reinforce, Rpt. ⫽ repeat.

exchanges used by the 14 teachers as they taught across the two subjects. Teacher explanation (teacher inform) and teacher-directed question-and-answer (teacher elicit and reinitiation) make up the majority of discourse moves in all 14 lessons, accounting for 78% of the total teaching exchanges in literacy and 77% of the total teaching exchanges in numeracy. The discourse analysis suggests teacher explanation was more common in literacy lessons; teacher questions and teacher direction (teacher directs) were more common in numeracy lessons. The discourse analysis revealed the rapid pace of teachers’ questioning and the predictable sequence of teacher-led recitation in which the parts are nearly always being played out as teacher–pupil–teacher. Pupil’s responses were often evaluated and commented on by the teacher, who maintained the right to determine what was relevant within her pedagogic agenda. Teacher-directed interrogation of the pupils’ knowledge and understanding was therefore the most common form of teacher/pupil interaction, with teacher questioning rarely going beyond the recall and clarification of information. Teachers therefore exercised close control over the nature, pace and direction of the knowledge pursued in the lessons. The findings therefore support the computerized systematic observation data, which suggest that the teaching was mainly interrogative and directive in nature. However, the aggregation of the discourse analysis data masks some of the individual variation found in the transcripts. Some of the teachers encouraged higher levels of pupil participation and engagement through open questions and different use of the follow-up move. Through feedback which went beyond evaluation of the pupil’s answer (i.e. probing and the use of uptake), teachers sometimes extended the answer to draw out its significance, or to make connections with other contributions during the lesson topic so as to encourage greater pupil participation.

408

F. Smith et al.

The length of pupil utterances was also analysed to explore to what extent pupils were encouraged to elaborate on their answers. Overall, we found that pupils were providing answers which were three words or fewer for 70% of the time. Pupil contributions were therefore rarely sustained or extended to encourage higher cognitive interactions. Comparison of the KS1 and KS2 teachers, however, showed some variation in the type of questions asked and the length of pupil utterance. KS1 teachers asked fewer open-ended questions (an average of 10% of all questions asked) compared to KS2 teachers (an average of 15%) and only 75% of responses were more than three words long. Therefore the transcript analysis suggests that KS1 teachers tended to use fewer challenging questions, which in turn encouraged fewer sustained responses. Discussion The article set out to investigate the impact of the official endorsement of ‘interactive whole class teaching’ on the interaction and discourse styles of primary teachers while teaching the NLS and NNS. The findings suggest that traditional patterns of whole class interaction have not been dramatically transformed by the strategies, supporting earlier studies of the NLS (Mroz et al., 2000; English et al., 2002; Hardman et al., 2003b). In the whole class section of literacy and numeracy lessons, teachers spent the majority of their time either explaining or using highly structured question and answer sequences. Far from encouraging and extending pupil contributions to promote higher levels of interaction and cognitive engagement, most of the questions asked were of a low cognitive level designed to funnel pupils’ response towards a required answer. Open questions made up 10% of the questioning exchanges and 15% of the sample did not ask any such questions. Probing by the teacher, where the teacher stayed with the same child to ask further questions to encourage sustained and extended dialogue, occurred in just over 11% of the questioning exchanges. Uptake questions occurred in only 4% of the teaching exchanges and 43% of the teachers did not use any such moves. Only rarely were teachers’ questions used to assist pupils to more complete or elaborated ideas. Most of the pupils’ exchanges were very short, with answers lasting on average 5 seconds, and were limited to three words or fewer for 70% of the time. It was also very rare for pupils to initiate the questioning. Effective teachers appeared to have a more interactive style as measured by the overall rate of discourse moves: on average they used 13% more discourse moves than the rest of the sample. However, the difference is a quantitative rather than a qualitative one: the types of discourse move did not differ significantly between the two groups of teachers. Compared to KS2 teachers, Reception and KS1 teachers used more directives and refocusing moves in their teaching and asked more closed questions. Choral responses were also more common among Reception and KS1 pupils and their answers tended to be shorter: three words or less for 80% of the time. Our study therefore supports the findings of English et al. (2002; p. 24), who found that ‘KS1 teachers tended to use higher levels of low cognitive interaction, fewer challenging questions and had fewer sustained interactions’. Differences in the

Interactive whole class teaching

409

teaching of literacy and numeracy also emerged: there was more teacher direction in numeracy lessons and closed questions and choral responses were more common than in literacy lessons. Explaining, uptake questions and open questions were also more likely to occur in literacy lessons. The questionnaire (see Hardman et al. [2003a] for a full discussion of the questionnaire) revealed that teachers had no clear concept of what interactive whole class teaching is, or shared language to discuss it, like English et al.’s study (2002). It also suggested that teachers had been given little practical guidance on how to implement interactive whole class teaching in the classroom. While over 70% of the sample indicated that they had received some training in interactive whole class teaching, less than a fifth of teachers had seen training materials published by the DfEE (1999b, c) and no one indicated that they had incorporated the ideas into their classroom practice. The questionnaire also revealed a mismatch between teachers’ perceptions of how they teach literacy and numeracy and the findings of the classroom observations. In response to questions exploring the frequency of use of strategies designed to promote pupil involvement in classroom discourse, most teachers reported that they valued and frequently invited pupils to elaborate on their answers. Our analysis suggests, however, that opportunities for sustained and extended dialogue by the pupil are rare. Overall, our findings suggest new ‘top–down’ curriculum initiatives like the NLS and NNS, while bringing about a scenario of change in curriculum design, often leave deeper levels of pedagogy untouched. Traditional patterns of whole class interaction persist, with teacher questioning only rarely being used to assist pupils to articulate more complete or elaborated ideas as recommended by the strategies. The findings also raise questions about the effectiveness of the in-service training programmes which have accompanied the national strategies. As Alexander (2000) argues, they point to the need for different approaches in order to change habitual classroom behaviours and traditional discourse patterns. The final evaluation of the NLS and NNS (Earl et al., 2003) suggests that changing such pedagogic understanding and practices remains a major challenge in securing the long-term effectiveness of the strategies. In order to bring about changes in the way teachers interact with their pupils, the findings suggest monitoring and self-evaluation need to become a regular part of inservice training, thereby giving teachers a degree of ownership of the process of school improvement (Mroz et al., 2000; English et al., 2002; Hardman et al., 2003b). Similarly, Joyce (1992) argues that teachers need extended opportunities to think through new ideas and to try out new practices, ideally in a context where they get feedback from a more expert practitioner and continue to refine their practice in collaboration with colleagues. Coaching and talk-analysis feedback may be useful tools for professional development whereby sympathetic discussion by groups of teachers of data (recordings and transcriptions) derived from their own classrooms could be an effective starting point for critical reflection. Such an approach could provide supportive interactions with peers through modelling and feedback in order to change traditional patterns of whole class interaction necessary for responsive teaching.

410

F. Smith et al.

The findings suggest the need for further research into ways of effectively supporting teachers in their professional development in order to promote more reciprocal forms of teaching to increase the opportunities for extended interactions with pupils. There is also a need for more research to provide comprehensive evidence, for both teachers and policy-makers, that interactive styles of teaching encouraging more active pupil involvement can produce significant gains in learning. Acknowledgement The authors wish to thank the Economic and Social Research Council for funding the research described in this article (Project No: R000239213). References Alexander, R. J. (2000) Culture and pedagogy: international comparisons in primary education (Oxford, Blackwell). Alexander, R. J., Willcocks, J. & Nelson, N. (1996) Discourse, pegagogy and the National Curriculum: change and continuity in primary schools, Research Papers in Education, 11(1), 81–120. Brown, M., Askew, M., Baker, D., Denvir, H. & Millet, A. (1998) Is the National Numeracy Strategy research-based? British Journal of Educational Studies, 46(4), 362–385. Department of Education and Employment (DfEE) (1998) The National Literacy Strategy: a framework for teaching (London, DfEE). Department of Education and Employment (DfEE) (1999a) The National Numeracy Strategy (London, DfEE). Department of Education and Employment (DfEE) (1999b) Talking in class (London, DfEE). Department of Education and Employment (DfEE) (1999c) Engaging all pupils (London, DfEE). Earl, L., Watson, N., Levin, B., et al. (2003) Watching and Learning. Final Report of the External Evaluation of England’s National Literacy and Numeracy Strategies (Ontario, Ontario Institute for Studies in Education, University of Toronto). Edwards, A. D. & Westgate, D. P. G. (1994) Investigating classroom talk (2nd edn) (London, Falmer Press). English, E., Hargreaves, L. & Hislam, J. (2002) Pedagogical dilemmas in the National Literacy Strategy: primary teachers’ perceptions, reflections and classroom behaviour, Cambridge Journal of Education, 32(1), 9–26. Galton, M., Hargreaves, L., Comber, C., Wall, D. & Pell, A. (1999) Inside the primary classroom: 20 years on (London, Routledge). Hardman, F., Smith, F., Wall, K. & Mroz, M. (2003a) Interactive whole class teaching in the literacy and numeracy lessons. A report for the Economic and Social Research Council, (Newcastle, University of Newcastle upon Tyne). Hardman, F., Smith, F. & Wall, K. (2003b) ‘Interactive whole class teaching’ in the National Literacy Strategy, Cambridge Journal of Education, 33(2). Joyce, B. (1992) Cooperative learning and staff development: teaching the method with the method, Cooperative Learning, 12(2), 10–13. Mehan, N. (1979) Learning lessons (Cambridge, MA, Harvard University Press). Mortimore, P., Sammons, P., Stoll, L., Lewis, D. & Ecob, R. (1988) School matters (Wells, Open Books). Mroz, M., Smith, F. & Hardman, F. (2000) The discourse of the Literacy Hour, Cambridge Journal of Education, 30(3), 379–390. Noldus Information Technology (1995) The Observer, Base Package for Windows, Reference manual, Version 3.0 edition (Wageningen, The Noldus Information Technology).

Interactive whole class teaching

411

Pollard, A., Broadfoot, P., Croll, P., Osborn, N. & Abbot, D. (1994) Changing English primary schools? (London, Cassell). Reynolds, D. (1998) Schooling for literacy: a review of research on teacher effectiveness and school effectiveness and its implications for contemporary educational policy, Educational Review, 50(2), 147–162. Reynolds, D. & Farrell, S. (1996) Worlds apart? A review of international studies of educational achievement involving England (London, HMSO). Reynolds, D. & Muus, D. (1999) The effective teaching of mathematics: a review of research, School Leadership & Management, 19(3), 273–288. Sinclair, J. & Coulthard, M. (1975) Towards an analysis of discourse: the English used by teachers and pupils (London, Oxford University Press). Smith, F. & Hardman, F. (2003) Using computerised observation as a tool for capturing classroom interaction, Educational Studies, 29(1), 39–47. Tharp, R. G. & Gallimore, R. (1988) Rousing minds to life: teaching, learning, and schooling in social context (Cambridge, Cambridge University Press). Tymms, P. (1999) Baseline assessment and monitoring in primary schools: achievements, attitudes, and value-added indicators (London, David Fulton).