PublicationsSystematic reviewsSEN & teacher/pupil interactionSEN & teacher/parent interaction - summary
A systematic review of interactions in pedagogical approaches with reported outcomes for the academic and social inclusion of pupils with special educational needs. Summary


The growing demand for inclusive practices within mainstream schools has resulted in classroom teachers having to take direct responsibility for the individual learning needs of all pupils within the setting, and reduced the expectation that support staff should be the primary practitioners for children with special educational needs (SEN). The belief in a need for special pedagogical approaches for these children has also been widely critiqued (e.g. Norwich and Lewis, 2001; Hart, 1996) and there has been a growing focus upon the teaching practices that can be, and are, more broadly used by mainstream practitioners. Central to all these approaches are the interactions that both create the learning context and operate within it.

Since The Warnock Report (DES, 1978), there has been increasing emphasis in England and Wales on the importance of including pupils with SEN in mainstream settings. The Special Educational Needs and Disability Act 2001 (SENDA) changed the legal rights of young people with disabilities and their parents, extending disability antidiscrimination legislation to schools. To overcome the opportunity for uncertainty in providing support appropriate to children with special educational needs, there is a Special Educational Needs Code of Practice (DENI, 1996; DfES, 2001; National Assembly Wales, 2004), which gives guidance to local education authorities (LEAs), governing bodies and schools. In addition, The National Curriculum Inclusion Statement (DfEE/QCA, 1999), to which all teachers in England must adhere, places a statutory requirement on mainstream schools to provide ‘effective learning opportunities for all pupils’ and sets out three ‘key principles for inclusion, requiring them to provide suitable learning challenges, to respond to pupils’ diverse learning needs, and to overcome potential barriers to learning and assessment for individuals and groups of learners. A recent OFSTED report (2004) found, however, that many schools still do not see themselves as having the skills, experience or resources to provide effectively for children with special educational needs. This is despite evidence that increasing numbers of children with SEN are making good progress.

This review follows on from the fi rst review in 2004 (Nind et al., 2004) which sought to identify how pedagogical approaches can effectively include children with SEN in mainstream classrooms. The 2004 review identifi ed a small evidence base to suggest that peer group interactive approaches were effective for the inclusion of children with special educational needs in mainstream classrooms, both in terms of social and academic participation. The study also identified the importance of the co-construction of knowledge through participation in the classroom learning community.  In addition to the 2004 review, there have been reviews that were technically non-systematic, which sought  to establish the effectiveness of particular pedagogies (Norwich and Lewis, 2001) or looked at approaches beyond classroom pedagogy (Sebba and Sachdev, 1997), and a number of systematic reviews that have considered the impact of broader school actions on pupil participation (Dyson et al., 2002; Harden et al., 2003; Howes et al., 2003). It has been noted elsewhere (Skidmore, 2004) that there is a need to explore more fully the individual interactions within the classroom in relation to effective inclusion.


This is the second year of a three-year project that is focusing upon effective pedagogical approaches in use in mainstream classrooms with children with special educational needs, aged 7–14 years. This second review expands the focus of the previous year to investigate the nature of the interactions between teachers, support staff and pupils.

Review questions

Our overall review question for the three-year project is:

Q1 What pedagogical approaches can effectively include children with special educational needs in mainstream classrooms?

Our in-depth review in the second year focuses on the more specific question:

Q2 What is the nature of the interactions in pedagogical approaches with reported outcomes for the academic and social inclusion of pupils with special educational needs?


In the first and second years of this review the overall question (Q1) was identified by the Review Group and agreed with the Advisory Groups. Q1 was the guiding question in both years for the subsequent electronic search of databases. This electronic search was carried out using a variety of keyword terms, drawn from the educational terminology of different countries, and from the British Education Thesaurus. All studies identified through these searches were imported into EndNote bibliographic software, and then into the EPPI-Centre systems. The same keywords and databases were used in both the fi rst and second reviews.

In both the fi rst review and this review, the citations were screened by two independent screeners, with a sample being evaluated by the EPPI-Centre link-person for quality assurance. The citations were initially screened on the basis of their titles and abstracts. This screening involved the application of eight agreed inclusion / exclusion criteria, which defi ned the subsequent scope of the review. To be included, the studies had to focus on pupils aged 7–14, with special educational needs, in mainstream classrooms. They had to include pedagogical approaches, offer an indication of pupil outcomes, and be empirical (in that they  involved the collection of data). They also had to be written in English and published after 1994. A range of electronic databases and citation indexes were searched as well as a variety of internet sites. Following the screening process, copies of papers were sought and given a second more detailed reading, where again the inclusion/exclusion criteria were applied. In the current review, this second screening included the reading of papers that had not arrived in time for the cut-off date for the fi rst review. The equivalent cut-off date for document retrieval for the second year was 31 March 2005. This second reading also involved  two independent screeners, with quality assurance provided by the EPPI-Centre link person.

The papers that passed through this screening process were now keyworded using two sets of keywords.  The fi rst set used the EPPI-Centre (2002a) Core Keywording Strategy for education (Version 0.9.7), while the second set used a review-specifi c strategy designed by the research team. This second keywording strategy was initially designed in 2004 for the fi rst review, but was updated and expanded in this review. This keywording was applied by pairs of reviewers, working independently and then moderating their fi ndings. The process was once more sampled for quality-assurance purposes by the EPPI-Centre link person. This keywording process created a ‘descriptive map’ of the studies. This map offers an overview of the studies and the research within them, giving details of their aims, methodologies, interventions, theoretical orientation, outcomes, and so on. The keywording process did not assess the quality of the studies.

The full Review Group now had detailed discussions about the priorities for the in-depth focus in this review. Drawing on identifi ed needs of users, it was decided that a priority should be interactions that involved unsupported mainstream classroom teachers, and that these studies should focus on teaching and learning with outcomes for the academic achievement and social inclusion of pupils with SEN, as these are priorities both for individual teachers and the schooling system as a whole. It was also decided that we would not focus on programmatic interactions nor studies that merely described classroom practices, without some form of evaluation or exploration of the variables within the setting. These priorities were transposed into new inclusion and exclusion criteria, and applied to the studies in the descriptive map so as  to produce the relevant studies for the in-depth review.

The studies identifi ed for the in-depth review were now closely assessed by two independent reviewers. Data-extraction was carried out using generic EPPI-Centre guidelines for education and review-specific  guidelines created by the Review Group, and any differences between the two reviewers were discussed and resolved. A central component of the two sets of dataextraction guidelines was the assessment of the quality of studies and weight of evidence (WoE) supplied by their fi ndings. Within the EPPI-Centre systematic review process, this is a key component in identifying the reliability and quality of each study, the trustworthiness of study results, and the weight of evidence that the study could contribute to answering the in-depth review questions. The reviewers assessed the relative WoE in relation to: the soundness of studies (internal methodological coherence, WoE A); the appropriateness of the research design and analysis in relation to the review questions (WoE B); and the relevance of the study topic focus to the review questions (WoE C). An overall weight of evidence valuation was arrived at through the combination  of weightings identifi ed in relation to the quality of execution, appropriateness of design and relevance of focus (WoE D).

The assessments of the reviewers were now used by the main authors to frame the synthesis of the studies, and the subsequent conclusions and recommendations. An evaluation of the quantitative and qualitative components of the studies was undertaken, identifying central themes and fi ndings across the studies, so that a structured narrative could be created which presented key aspects of interactions in effective pedagogical approaches.


A Across the two years, 3,324 papers were identified for potential inclusion. Having excluded duplicates, 2,812 were initially screened on the basis of their titles and abstracts or by hand, and 2,224 were excluded  for not meeting the inclusion / exclusion criteria of the review. Of these the most common reasons for exclusion were not being empirical studies (30%), not being concerned with pedagogical approaches (29%), and not indicating pupil outcomes (22%). This meant that 587 papers were sought in 2004–2005 to have the inclusion / exclusion criteria applied during a more detailed reading. 70 papers were not obtained by the cut-off date. 517 full documents were screened, with 405 papers excluded. In the application of exclusion / inclusion criteria to the collection of titles and abstracts, the measure of inter-rater reliability between the two members of the Review Group was good in both years (Cohen’s Kappa 2004:0.62; 2005: 0.65). Again the three most common criteria for exclusion were the categories identifi ed above. Four studies were also found to be reported in two papers. The systematic map therefore included 109 studies (68 from 2004 and 31 from 2005).

91% of the studies were identifi ed through electronic databases, 83% came from the USA and 9% of the studies came from the UK. Over 90% of the studies were either evaluations or explorations of elationships,  and over 80% focused upon Teaching and Learning. 55% of the studies claimed an impact upon academic attainment and 44% upon social interaction/involvement. Only 31 studies (28%) focused upon the regular teacher working on their own in classroom, yet the majority of studies gave some evidence about pupil-teacher interactions (83%) and far less about the interactions involving support staff (for example, pupils-support staff interactions: 18%). The majority of these interactions were informal (72%) and considered (68%), with the minority being to some degree programmed in nature (26%). Particularly noticeable too was the emphasis upon verbal (84%) and written (64%) interactions, in comparison to other  forms, particularly tactile (15%) and signed (1%) interactions.

Seven studies met the criteria for inclusion and were included in the in-depth review. They cover a range of settings, subject areas and research types. Five of the studies are from the USA and one each from Canada and Australia. The studies are equally divided between primary and secondary phases of education, and while three were conducted within science classes, two do not have a specifi c curricular focus, one draws upon a general curriculum, and the other upon literacy. There was a broad mix of special
educational needs focused upon within the studies, including those with learning impairments, physical impairments, sensory impairments, and emotional and behavioural diffi culties. Four of the studies have verbal interactions to the fore, with written, technological and auditory interactions being considered in the other papers. Five studies evaluated settings without researcher-manipulation (N=2) or with researcher-manipulation (N=3) and two studies primarily explored the relationships between variables within the setting.

Synthesis of these studies lead to the following four major themes emerging:

  • interaction and the mediating role of the teacher
  • interaction, cognitive level and engagement
  • interaction and the learner’s voice
  • interaction and knowledge as contextuallygrounded

Weight of evidence (WoE)

None of the seven studies was allocated a high WoE in relation to its trustworthiness in answering its own study question (WoE A);, fi ve received a medium rating; and two were allocated a low rating. In relation to the specifi c question of the systematic review, we considered the appropriateness of each study’s research design and analysis (WoE B), and also considered its relevance (WoE C). Two papers (Palincsar  et al., 2001; Wallace et al., 2002) were deemed to be of high trustworthiness in relation to WoE B and C;  two further studies (Jordan and Stanovich, 2001; Rieth et al., 2003) were allocated a medium rating on both these criteria; another (Tindal and Nolet, 1996) obtained a medium rating for WoE B but a low for WoE C, while the two remaining studies (Ward and Center, 1999; Zembylas and Isenbarger, 2002) scored low on both criteria. In terms of overall weight of evidence (WoE D) a majority of studies (5) were  deemed to be medium in trustworthiness and a minority (2) were deemed low.


In synthesising our fi ndings, our conclusions refl ect the WoE which we can apply to the studies. A common theme across all the studies is the powerful role the teacher plays in shaping interactions and influencing learning opportunities through interactions. Six of the studies observed teacher interactions while  the one remaining study audio-recorded classroom interactions involving a direct focus on teachers’ interactions. It is evident that positive teacher attitudes towards the inclusion of children with special educational needs are refl ected in the quality of their interactional patterns with all pupils and, in turn, to their pupils’ self-concept. Those teachers who see themselves responsible for fostering the learning of all promoted higher order interaction and engaged in prolonged interactions with pupils with special educational needs, while teachers who see others (e.g. specialist teachers or special education teachers) as primarily responsible for these pupils engaged in interactions that were of a non-academic and low level nature.

Those interactions that are demonstrated to be more successful in terms of academic and social outcomes are characterised by questions and statements involving higher order thinking, reasoning, and implicating a point of view. The teachers who enable pupils to achieve these outcomes were those who spent most of the available time in these high-quality, on-task interactions as opposed to the low-quality off-task interactions. High-quality interactions are those in which teachers offer learners the opportunity to problem-solve, to discuss and describe their ideas, and to make connections with their own experiences and prior understandings, while those teacher interactions that are less successful focus on procedural matters, behaviours and general classroom management.

The theme of the learner’s voice emerges explicitly from fi ve of the seven studies in the in-depth review. Pupils with special educational needs participated more fully when encouraged to identify their thoughts and assisted to document them, particularly through one-to-one discussion with the teacher. The importance of the teacher eliciting prior knowledge and understanding was also evident, and, in two studies, it was noted that this enquiry had resulted in teachers being impressed by the thinking and conceptual understanding of pupils with special educational needs. Successful interactions were also recognised as those in which the teacher calibrated questions and answers to the pupils’ responses, following the pupils’ thinking rather than just checking that their understanding equated with that of the teacher.

The importance of interactions being based in learners’ experiences is a theme emerging from three studies.  Drawing upon this contextuallygrounded knowledge builds connections to the authenticity of activities and the perception that they are meaningful to learners in the here and now of their lives. These interactions involve direct experiences and realistic problems, offering multiple opportunities to engage with the learning situation and others within it. The high level of higher order thinking in these approaches suggests that contextualising what is to be learned in the form of inquiring into tangible problems has potential to foster academic and social inclusion of pupils with SEN.

Strengths and limitations

This systematic literature review had both strengths and limitations. The literature review was limited in scope to material from 1994 and to pupils aged between 7 and 14, but it drew on evidence from a full range of pupils and settings in this age group. It included studies that represented a broad range of SEN, and offered a reasonable range of curricula foci. It also drew upon studies of varying size, from a case-study of one child to a study of 118 classrooms, although the number of studies was small. The review benefi ted from highquality assurance, with screening, data-extraction and quality-assessment being conducted by two independent review team members (or a Review Group member and EPPI-Centre link person) at each stage. The quality of the studies within the review and the rigorous check on quality further
strengthens confi dence in the review fi ndings.

The number of studies that did not arrive in time (70 out of a possible 587, 12%) is a potential limitation of the review, as is the comparative lack of studies that presented negative or null outcomes. There is also a total lack of studies originating in the UK, which limits certainty about the context and cultural equivalence of studies, and therefore the generalisability of fi ndings, although all studies were conducted in English language settings. Although the majority of studies were allocated a medium weight of evidence rating overall (WoE D), the absence of studies with a high rating overall is another limiting factor that must be taken into account.


Researchers, policy-makers and practitioners should be aware that there is a shortage of evidence about the nature of teaching approaches that effectively include children with SEN in mainstream classrooms. In addition there is a shortage of evidence about teachers working alone within inclusive settings, and about their interactions with pupils, particularly in relation to interactions involving tactile and signed modes of

There is evidence, however, particularly in relation to oral interactions, that teachers are more likely to be effective with all pupils if they use language to draw out pupils’ understanding, and encourage further questioning and links between new and prior knowledge. These interactions are more likely to be effective  if they are situated within activities that are hands-on, personally relevant and offer a range of opportunities  to engage with the concepts, and with others’ understandings of those concepts.

Given the complexities of working within inclusive settings, teachers in training need opportunities to reflect  on their practices in the light of the existing research base. The fi ndings of this review underline the importance of this in particular, since it strongly supports the notion that teachers who see the inclusion of  pupils with SEN as part of their role are more likely to have effective, high-quality, ontask interactions, and less likely to focus on relatively ineffective organisational and behavioural matters when talking to pupils.

This report should be cited as: Rix J, Hall K, Nind M, Sheehy K, Wearmouth J (2006) A systematic review of interactions in pedagogical approaches with reported outcomes for the academic and social inclusion of pupils with special educational needs. Technical report. In: Research Evidence in Education Library. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London.

Copyright 2019 Social Science Research Unit, UCL Institute of Education :: Privacy Statement :: Terms Of Use :: Site Map :: Login
About::Projects::Publications::Research Use::Courses & seminars::Resources::Databases::Blog