Towards an evidence-based approach to fostering collaborative conversation in mainstream primary classrooms: Response to commentators

Abstract

The ability to engage with ease in collaborative conversation is critical for child well-being and development. While key underpinning skills are biologically enabled, children require appropriate scaffolding and practice opportunities to develop proficient social conversational ability. Teaching conversation skills is a statutory requirement of the English primary (and many other) curricula. However, currently most upper primary mainstream teachers are not trained to teach conversation skills and do not teach them in the classroom or provide time for children to practice. We argue for first steps towards an evidence-based approach for a universal/Tier 1 programme, while fully acknowledging an ongoing need for Tier 2 and Tier 3 support as well as for further research into the strategies which are most effective in Tier 2/3 contexts. Further research is also needed to explore cultural variation in social conversation and to develop reliable, valid and brief teacher measures of child social conversational ability.

Importance of social conversation

There is a general consensus in Bryant, Hindman and Chor, and Snow et al.'s responses to our review article (Abbot-Smith et al., 2023). We all agree that children’s social conversational ability is critical for their well-being and development and that extra steps should be implemented to support it during childhood. Nevertheless, there is some debate about how best to do so, and there remain many unresolved empirical questions.

Cultural variation and which skills to target

Bryant (2023) raises important issues about, first, cultural variation in how adults engage in social conversation as well as, second, the challenges in knowing which conversation skills to prioritise. Both impact on how conversational skills are supported. With regard to cross-cultural differences, stringent comparisons in relation to conversation are scarce (cf. Stivers et al., 2009). There are also likely to be some approximate universals underpinned by the cooperative nature of conversation. It should therefore be possible to support the development – and meta-linguistic awareness – of such things as turn-taking, active listening and responding, perspective-taking and making a relevant response to the immediately preceding conversational turn. Several aspects are likely to be challenging for many children given the speed and demands of conversation. This is the kind of ‘cultural and social capital’ to which all children should have access in order to navigate a conversation and share the load of repairing breakdowns.

Combining explicit teaching and peer practice versus scaffolding, modelling and providing conversational opportunities in the classroom

Opinions differed among the commentators regarding whether mainstream classrooms should include explicit highlighting, discussion and teaching of social conversation skills. The nature of appropriate support is likely to vary according to children’s developmental stage. We agree with Hindman and Chor (2023) that implicitly teaching children how to talk, together with feedback on child-to-child talk which is elicited through ‘strategically co-ordinated books, hand-on manipulatives and activities’, may well be preferable prior to 6 years. It is less clear how to embed these skills in classroom contexts for older children. Our focus is to promote social conversation in older primary years groups (ages 7 through to 11 years), where currently there is very little opportunity for ‘free’ conversation in the vast majority of classrooms and – as a result – limited opportunity to further nurture conversational skills that still have much room to develop.

Social conversational ability as a ‘biologically primary skill’ which ‘does not need to be taught’

As Snow et al. (2023) rightly point out, social conversation and its prerequisite skills (oral language ability, affect recognition, mentalising) are observed universally in human cultures. This could imply that social conversation does not need to be explicitly taught. While we appreciate the distinction being made here and agree that the basic mechanics of conversation (e.g. knowing that a response should be relevant) may well be biologically primary, it does not follow that becoming proficient in social conversations will develop without appropriate scaffolding and opportunities.

Moreover, there is significant variability even among neurotypical adults with regard to excessive verbosity/reticence as well as in the use of verbal and non-verbal behaviours which indicate active listening. These individual differences are linked to social desirability ratings and relationship success in adults (Arkowitz et al., 1975; Miczo et al., 2001; Wheeless et al., 1992). Therefore, it seems there may well be a case for Tier 1 (i.e. universal) teaching although we take Snow et al.’s point that children also need to learn many other things in school.

Focussing on teacher screening and Tier 2 intervention

Instead of aiming to improve social conversation skills as part of Tier 1 provision, Snow et al. (2023) argue that teachers should be trained to accurately use language and communication screening tools and that adequate Tier 2 (small group) and Tier 3 (one to one) language and social communication therapy should be provided for children who score below a certain threshold (Ebbels et al., 2019). However, models of intervention based on response to intervention are premised on quality first Tier 1 instruction, in this case related to conversational skills. The current evidence suggests this is lacking in primary schools and teachers are not trained to determine which children need additional oral language support in later primary years (Dockrell & Hurry, 2018).

Indeed, a recurring point, critical for all approaches, is that there should be a greater emphasis on training teachers to understand oral language development, both in their initial training and in continuing professional development. To date, evidence in this area is mixed (Goldfeld et al., 2022; Thurston et al., 2016). Thus, it may well be that in addition to increased knowledge, teaching staff require support and feedback when putting their knowledge into practice.

Aside from the issue of teacher training, it is worth reiterating that if conversation support is restricted solely to those children whose ability is markedly impaired, there are a great deal of children who, although not reaching the threshold of a diagnosable condition, will miss out on support that could have consequences both in the classroom and beyond (Donno et al., 2010; Helland et al., 2014). Finally, since conversation is a cooperative endeavour, there is a case to be made for raising meta-linguistic insight across the board such that those who are more able can appreciate diversity and share the load with children who struggle for other reasons (e.g. due to speed of language processing, hearing, attention). Of course, it is an empirical question whether this is so, and a greater evidence base is certainly needed before implementing change in Tier 1 time.

Assessing improvement in conversational ability

Whether the way forward is via Tier 2/3 interventions or (as we argue) through a combination of improved Tier 1 in addition to Tier 2/3 intervention, Bryant (2023) notes quite rightly that it is currently difficult to know which measures to use to determine intervention or programme success (see Timler & Covey, 2021 for discussion of problematic aspects of existing standardised direct measures of pragmatics).

While we agree in principle with her suggestion of combining direct video-recorded measures with teacher assessments, there are currently issues with both. Video-recorded measures can be used by trained researchers (Dillon et al., 2021; Gangi et al., 2021; Nadig et al., 2010; Spitzberg & Adams, 2007) but they are too time-consuming for use by teachers. Standardised questionnaire measures are very broad and do not capture many aspects of conversational ability (Bishop, 2003; Gardner & Curenton, 2017; Gentilleau-Lambin et al., 2019). One crucial target for future research should therefore be to develop and standardise a relatively quick questionnaire of conversation skills designed for use by teachers.

Emergence through semi-naturalistic scaffolding alone? Social conversation targets on national curricula

We agree with Snow et al.’s point that ‘in reality, adults (parents and teachers in particular) invest considerable time into scaffolding the emergence of conversational discourse skills through real-time feedback and modelling’. Moreover, over the last decade, the evidence for the importance of child exposure to conversational turn-taking has become particularly clear (Donnelly & Kidd, 2021; Zimmerman et al., 2009). Unfortunately, a substantial proportion of neurotypical children may experience limited conversational turn-taking in their home environments for a plethora of reasons. The school environment does have the potential to ‘level the playing field’ in this regard.

While we acknowledge that teachers cannot be expected to do everything, Bryant (2023) highlights our original point that conversation skills are already explicitly a statutory requirement on many educational curricula. In the primary curriculum for spoken English, the UK Department for Education (2015) lists as a statutory requirement that ‘pupils [in years 1-6] should be taught to:

• Listen and respond appropriately to adults and their peers

• Maintain attention and participate actively in collaborative conversation, staying on topic and initiating and responding to comments

• Gain, maintain and monitor the interest of the listener(s)’.

It is of course entirely possible to teach listening and responding appropriately to adults and peers within a whole classroom context of discussion around topics such as geography, history, science. However, the reality is that most teachers, even for the ‘early years’ (in the United Kingdom up until age 6 years), provide very little opportunity for any kind of conversation, as noted by Hindman and Chor (see, for example, Wasik et al., 2022). Opportunities for practising oral language fluency of any kind – let alone social conversation – within the classroom itself become even more limited from 7 years and up (Alexander, 2013). Rather, oral language contributions tend to be controlled by and dominated by the teacher (Mercer, 1995).

Teachers are too busy and the curriculum is too full: teachers’ own views on what is feasible and acceptable

We shared some of the concerns outlined in Snow et al.’s and Hindman and Chor’s responses that primary school teachers could struggle to fit a large number of explicit lessons on conversation skills into their packed timetables. Hindman and Chor’s point about workforce stabilisation is also critical. To explore teachers’ perspectives further, we sought the views of UK teachers themselves, first via a survey (of 79 teachers), followed up with interviews (N = 10) and a focus group (N = 8) consisting of upper primary teachers (Sturrock et al., under review).

Initially, when presented with the above curriculum targets, only 6% of participants spontaneously mentioned said they explicitly taught collaborative conversation or its prerequisite skills. After watching our brief Continuing Professional Development video, the vast majority said they felt willing and able to teach and support social conversation with the following caveat. To fit this in, teachers proposed to have pre-made materials to involve many short activities which could be slotted in briefly but frequently and could thus also maximise pockets of time which are hard to plan for and so often ‘go wasted’. Another proposed solution was for activities to be designed to fit into other curriculum activities, such as (in the UK curriculum) Personal Social Health and Economics (PSHE). ¹ A final suggestion is that the core skills of active listening, monitoring listener interest, turn-taking and providing relevant responses are reinforced in paired and small group discussion of other curriculum content (Hindman et al., 2022).

Where to go from here

In sum, there is much consensus. We all agree that we need further research to better understand how to support conversational development and how to best target support. In an effort to determine the utility of a whole classroom approach, we are co-designing materials and procedures with UK upper primary teachers and will run a preliminary feasibility and acceptability study (https://www.nuffieldfoundation.org/project/classroom-intervention-to-improve-conversation-skills). For a universal approach to be truly universal within countries like the United Kingdom, United States, Australia and New Zealand, it is imperative that it be appropriate for children from a range of cultural backgrounds, which means that basic research in cross-cultural comparisons is also required. Further research is also needed in relation to Tier 2/3 interventions. All of these approaches also require further developments in valid, reliable and easy-to-administer measures of child conversation.

Another reoccurring theme throughout the commentaries is the central place of the teacher and, therefore, the imperative for an increased focus on teacher understanding of child oral language development as part of initial training and as part of their continuing professional development. None of the suggestions raised either by us or by the commentaries can be successful if teachers do not have the motivation and understanding to prioritise child oral language and social communication development. Hindman and Chor are entirely right to highlight that a fundamental first step is for teachers (and early years professionals) to receive adequate renumeration, respect and support to ensure staff retention and to allow the headspace to support research and engage with its outcomes.

Most importantly, we all agree that children with weak social conversation skills are sorely disadvantaged. Supporting this core developmental ability should support peer relations and all the benefits which stem from them.