Group discussion and extended writing

by Dave Spence, Wilmslow High School 

Project rationale 

An analysis of KS4 data revealed that students were underperforming in extended written responses on English literature texts in relation to the national average. In particular, boys were struggling to produce substantive written responses in exam conditions and a further sub-group of disadvantaged students were attaining well below their target grades. 

Having seen the impact of oracy strategies on improving KS2 and KS3 students’ written work, I wanted to investigate whether Harkness-style discussions would improve older students’ written work. Initially, this involved group discussions on an aspect of the literature text being studied without prior preparation, but as this appeared to have little impact on improving written responses, I decided to explore how students responded when coming into discussions having completed prior research. An example of this would be a student joining a group discussion on A Christmas Carol having prepared some research or answers to questions on one specific character while their peers prepared responses on other characters or themes in the novel. 

The group I decided to focus on was a Year 11 mixed ability group focusing, as part of their revision, on producing regular timed responses to texts for the literature examinations. The research question I investigated was: to what extend does pre-prepared group discussion impact on students’ extended written assessments? 

Baseline data

Intervention and intended impact

The intention was to follow the RDW (read-discuss-write) model, whereby students undertook independent research before a group discussion on a topic. I was concerned with the possibility that, for some students, the opportunity to ‘piggyback’ off other students’ work in discussions, as Lemov (2015) outlines, could distort the impact of this approach. 

The likelihood of acquiring a ‘false positive’ in relation to marks of students’ subsequent written assessments was one that I felt I could not really guard against in this project. The idea of students using their ‘recollection of the discussion to write, potentially crowding out the need to learn from other sources of knowledge, such as reading’ (Lemov, 2015) was one that I was also very conscious of and concerned about, in relation to the long term effects of such strategies. 

However, even if students were using the ideas from discussion to conceal a lack of depth of knowledge, would this be such a bad thing? The aim was to encourage extended responses and if this was achieved through relatively superficial knowledge acquisition, this could be considered to be an acceptable pay-off at this point. ‘Build Stamina’, advocated by Lemov (2015) was introduced to support students to develop confidence in expressing ideas on paper that they had previously tested in discussion. Daniel Willingham expresses this conundrum regarding the usefulness of talk in a discussion with Martin Robinson in Trivium 21c:

‘When kids have a lot of dialectic, but not much knowledge, the dialectic is not very effectively deployed. They have great fun talking with one another, but they’re really recreating the thinking of other people. It’s useful in its way, but it might be more useful if they just read up a little bit and then they would have a more advanced starting point.’ 

Willingham (2013)

In a mixed ability group and with careful engineering of the ‘talk groups’, it could be argued that certain higher attaining students would dominate the ideological approach in the discussions, but this did not preclude all students having the opportunity to engage in the process and offer their ideas, even if this was, to a certain extent regurgitating others’ ideas. This could be considered simple repetition, but it also evidenced listening skills, recall and application when it came to writing exam-style responses. 

The students who had historically struggled with producing extended written work were also those who have struggled with using and understanding ‘Tier 2 vocabulary’. This had been identified through numerous vocabulary tests in the first half term. . Having read about the important link between vocabulary knowledge and confident written expression in ‘Robust Vocabulary Instruction’ (2013) and ‘The Early Catastrophe’ (1995) alongside Alex Quigley’s work on ‘Closing the Vocabulary Gap’ (2018), I also wanted to explore how group discussions would help these students use more academic vocabulary and some Tier 3 subject specific language when developing their lines of argument.

When setting up the discussions, I emphasised the need for students to use high level academic vocabulary wherever possible and the vast majority of students came to the discussions having completed not just background research, but also vocabulary preparation and this was clear when filming the discussions.

Impact data

Assessment data from the sample group reveals that, on average, students attained 2.6 marks more on literature assessments following pre-prepared group discussions, compared with when they completed the assessments with normal independent ‘homework’ preparation.

An individual example which illustrates this improvment is Pupil 4 (an underachieving boy) who scored 10 and 9 marks in essays on A Christmas Carol and Power and Conflict Poetry respectively when asked to prepare for the assessment individually (identified as ‘no prep’ in the data table above). These results then improved to marks of 11, 12 and 13 following peer group discussions in advance of exam-condition assessments. 

A similar pattern of results was seen across the rest of the class as a whole. In particular, boys consistently performed better across the six assessments when they had had an opportunity to discuss their ideas with others beforehand. These findings were also evident with two SEN students whose attainment improved by up to six marks following these oracy based interventions.

Students were also asked to complete a student voice questionnaire anonymously at the end of the project, where they were asked to respond to a number of questions and submit a form online. Eight students were asked to complete the survey, and there was a 50% completion rate. The four students who completed the survey were unanimously positive about the intervention, noting that their confidence in speaking and listening had improved as a result. However, this may not be representative of the wider population, since students with less confidence may not have responded to the survey.


Getting students talking between themselves about a topic and moving away from an overly didactic ‘chalk and talk’ approach to teaching literature is an area that I feel can make a huge difference both in terms of engagement and attainment. 

The evidence from this small research project suggests that this approach could have a positive impact on student attainment if adopted more systematically. Students’ marks across all abilities went up following this oracy initiative, and this was particularly pronounced when looking at SEN students’ attainment.

From the data I collected, I cannot be sure whether this approach helped students engage with texts with greater enthusiasm. However, my sense from observing students during the group discussions is that the whole class were actively involved in learning, and delivering their points thoughtfully and with conviction. 

As Martin Robinson writes in Trivium 21:

“Children… need the opportunity to express themselves, whether they are right or wrong, in a framework that allows them to engage… it can drive the learning forward as students engage more fully with the learning. Our schools must resound to the sound of debate”. 

Robinson (2013)

Results of the students’ summative external assessments are yet to be seen and will be influenced by numerous other variables. However, this research project approach suggests that encouraging students to engage in research-informed discussions may produce sustained improvements to analytical writing. This is an area that I will continue to explore and promote with students in the future. 


Robinson, M. (2013) Trivium 21c: Preparing Young People for the Future with Lessons from the Past. Independent Thinking Press.

Lemov, D. (2015). Teach Like a Champion 2.0: 62 Techniques That Put Students on the Path to College. San Francisco, CA: Jossey-Bass. 

Hart B., Risley T., 1995, The early catastrophe: The 30 million words gap by age 3, American Educator, 27, 4-9.

Quigley, A. (2018) Closing the Vocabulary Gap. Routledge: London.

Beck, I., McKeown, M. (2002). Bringing words to life: Robust vocabulary instruction. New York: Guilford. 

Featured image: Pixabay

Leave a Reply