by Prof. Christine Howe, University of Cambridge
Introduction
Based on research conducted during the latter part of the twentieth century (e.g., Galton et al., 1999), group work amongst British primary school students is frequently portrayed in very negative terms. Students are depicted as seated in groups for classroom activities but seldom as working productively with their groups. Independent activity is seen as the norm even when teachers ask for collaboration, and dialogue is said to be low-level and often off-task. The possibility that students might learn from group work with classmates is widely regarded as unthinkable. Thus, when an article appeared in the 2018 Times Education Supplement entitled ‘Why it’s time to say goodbye to group work’ (Hallahan, 2018), it will undoubtedly have struck a chord.
With the Covid pandemic and consequent school closure, group work has largely been abandoned just now. However, the pandemic will not last forever, so should group work be resuscitated as schools get back to normal or should it be allowed to rest in peace? In this blog, I shall summarise analyses of a recently obtained and large-scale dataset, results of which show, firstly, that group work amongst students is both widespread across primary classrooms and also often of very high quality. They show, secondly, that the quality of group work has direct relevance to student attainment, including in high stakes public examinations. They show, thirdly, that group work can provide a platform for subsequent, highly productive interaction, including with teachers. Thus, I shall argue that far from celebrating group work’s current suspension, we should re-invigorate it as soon as we can, and never willingly say goodbye.
Group Work and Student Attainment
The dataset covers 72 primary school classrooms, situated in urban and rural locations across many parts of England (see Howe et al, 2019 and Howe, 2020 for detailed information about the dataset, its analysis and most of the results reported below). The classrooms were either Year 6 or Year 5/6 composites, but were demographically diverse (e.g., 0-100% of students per class eligible for free school meals; 0-96% from minority ethnic backgrounds). Video and observational data were obtained during literacy, mathematics or science lessons, with analyses based on two lessons per classroom (covering any pair of the three possible subjects). Video data focused on teacher-student interactions while observational data included ratings of group work, using a well-validated instrument derived from previous research (Kutnick & Blatchford, 2014; Howe et al, 2007).
In particular, whenever a teacher asked the students to work independently in small groups, a researcher chose one group at random and monitored their interaction. Employing time-sampling techniques (10-seconds per minute preparation, 20-seconds observation, 10-seconds recording on grids, 20-seconds rest), the researcher noted usage of such features as justified reasoning, elaboration of expressed ideas, productive approaches to disagreement, and striving for consensus. Noted features were then converted to ratings (1=Not true, 2=Partly true, 3=Very true) on five scales. Five further 3-point scales were used to rate more general aspects of group functioning, for instance on-task activity, positive attitude, and involvement of all group members. The researcher stayed with the same group over very short sessions of group work punctuated by interaction with teachers, and she applied the scales across all of these sessions. Otherwise, she moved to and rated different groups for each group-work session to maximise representativeness.
Group work was observed in 71 of the 72 classrooms, with between 0 and 8 rated sessions per lesson, each session lasting between 1 and 42 minutes, with a mean duration of 6.47 minutes. The ratings were regarded as collectively providing evidence about quality, for they reflect features that existing research associates with positive student outcomes (e.g., Kutnick & Blatchford, 2014; Howe & Mercer, 2007). Certainly, high scores would challenge the image sketched above of independent work, low-level dialogue, and off-task activity, and in general high scores were obtained. Analyses warranted averaging across all ratings obtained for each classroom, that is awarding each classroom a composite mean of between 1.00 and 3.00. The average of these composites across the 71 relevant classrooms was an impressive 2.26. All in all then, group work amongst students was indeed both widespread across this diverse sample of classrooms and of generally high quality.
Nevertheless, despite the high average, what became known as the Group Quality Index still varied across classrooms, from a minimum of 1.55 to a maximum of 2.74 – and this variation turned out to matter. Taking student academic attainment at the start of the school year into account (together with other potentially relevant factors), scores on the Index were found to be strongly related to end-of-year performance on both a test of general reasoning and all three nationwide Standard Assessment Tests (SATs). It was only with a science test that no relation was detected. Specifically, the higher score on the Index, the better the Year 6 students performed on the reasoning test and on the Maths SAT, the Reading SAT, and the SAT in spelling, punctuation and grammar (SPAG SAT). While previous research has demonstrated associations between student attainment and the features represented in the Index, the focus has hitherto been upon reasoning and/or specific aspects of the curriculum. The present analyses are amongst the very small number to examine and then demonstrate associations with high stakes, public examinations like SATs.
Beyond-Group Sharing
Over the years, greater understanding has been reached about the processes by which students learn from small group activity. These processes are no longer thought to depend upon progressive group solutions that individual students internalise, as say a classic Vygotskyan model implies [e.g., Vygotsky, 1998). Rather, they involve active reflection upon group activity, which can take place as much as several weeks later and bring in additional information that is imbued with significance through linkage with group work. It is immaterial whether group solutions are progressive, identical to members’ initial views, or even regressive. The implication is that in addition to supporting the features represented in the Group Quality Index (which relate to group interaction per se), support should also be given to the reflective processes on which progress depends. This implication was also explored with the dataset I am focusing upon [see Howe, 2020).
While there are several ways in which students could be encouraged to reflect upon group work, the focus of the analyses was ‘beyond-group sharing’, that is occasions where students outline aspects of their group interaction to classmates who had not participated. On average, beyond-group sharing was detected after 23% of the observed group sessions, with the frequency per classroom varying from 0% of group sessions to 86%. Sometimes, sharing involved students joining new groups and reporting on their previous group. For instance, in one class, groups discussed statements about the solar system like ‘The Earth travels round the Sun every 24 hours’, and decided which statement was correct. Subsequently, a ‘host’ member of each group remained stationary, and the other members joined a different host. These newly formed groups shared ideas from their previous group and reviewed their decisions. At other times, sharing involved plenaries where students reported their group’s thinking to the whole class, an example being when pairs of students prepared interviews in the style of television news, one student acting as the interviewer and the other as the interviewee, and the pairs later role-played their interviews in front of the class.
Analyses indicated that, with student start-of year performance and other relevant factors considered, classroom use of beyond-group sharing was positively associated with student performance on the reasoning test, the science test, and to a more modest degree SPAG SATs. Here the relation was independent of scores on the Group Quality Index: with these measures students whose group activity was relatively undistinguished were as likely to benefit from sharing as students whose group work was exemplary. Use of beyond-group sharing was also strongly and positively associated with performance on Reading SATs, but this time only when scores on the Group Quality Index were relatively high. Altogether, it was only Maths SAT that failed to show positive effects from beyond-group sharing, although as with Reading the picture varied with Group Quality scores.
Sharing and Teacher-Student Dialogue
Besides its inherent advantages, beyond-group sharing guarantees high levels of student participation almost by definition, and analyses that focused upon teacher-student dialogue rather than group work (see especially Howe et al, 2019) spotlighted student participation as relevant to attainment. Specifically, the results showed that, again with start-of-year attainment and other relevant factors considered (including quality of group work), high levels of participation were strongly and positively associated with student scores on SPAG and Maths SATs, so long as levels of teacher-student ‘elaboration’ and ‘querying’ were also high. Elaboration was defined as building on, elaborating, clarifying or evaluating an earlier contribution (or inviting any of these), while querying was defined as doubting, challenging, rejecting or showing disagreement.
Noting that student participation and elaboration/querying must operate in tandem for teacher-student dialogue to have positive effects and that beyond-group sharing is inherently participative, I wondered whether beyond-group sharing could act as a platform for elaboration and querying. Recent in-depth analysis of eight of the dataset’s classrooms indicates that it can do this, so long as four conditions are met:
- The whole-class plenary form of sharing is employed, since this facilitates teacher involvement and hence teacher-student dialogue.
- The task undertaken during the preceding group work guarantees a range of student ideas.
- Beyond-group sharing involves representatives from most (ideally all) groups reporting their group’s thoughts about the task, no matter whether those thoughts are consensual or divergent. The need for group thoughts must be emphasised: teachers often solicit personal conclusions after group work, usually to check on accuracy, but these are beside the point here.
- Students have a strong sense that through sharing a bigger picture is being created. Often this sense was created in the analysed classrooms through the teachers inviting ‘adding’ or ‘building’, but sometimes it was clear that the students were achieving it without explicit support. In any event, when the conditions were met, elaboration occurred with three or four times its average frequency across all classrooms within the dataset, and querying occurred with two or three times its average frequency.
The (lightly edited) extract that follows illustrates how teacher-student elaboration and querying abound during beyond-group sharing when the four conditions are met. The extract comprises the first half of roughly six minutes of sharing, which followed small group discussion of the statement ‘It was right for the dogs to be shot’ in the context of Shackleton’s expedition to the Antarctic. (All names are gender and ethnicity appropriate pseudonyms).
Teacher: Right, it’s time to summarise, get together the gist of your group, and then you need to report back. Right, so, nice and clearly, and take on a point of everybody in your group, and then we’ll see how the conversation went. OK, can we start with Saleem?
Saleem: In my group, Angela, Anna and Mysha all agreed with it was right to shoot the dogs because, like in the book, they said they needed more space on the ships and they needed more food supplies. Me and Mysha agreed that it was right, and we agreed and also disagreed with the statement, because they could have used the dogs for guard dogs or something like… transportation.
Teacher: OK… OK, right, thank you for your point. Who would like to add or build on what Saleem’s just said? Ayesha.
Ayesha: Building on what Saleem said, I disagree because how will the guard dogs protect the stuff, because, say, for example, some of the men, they were hunting and polar bears came, then the guard dogs wouldn’t be enough to guard their belongings.
Teacher: OK. But what was the overall view, opinion?
Ayesha: Overall, we think that it was right for the dogs to be shot because, if they don’t shoot the dogs, then the humans would starve, and the dogs would starve as well because there wouldn’t be any food left. So, if they shot the dogs, the humans would have the food. Priority is that the humans survive.
Teacher: Survive, OK. Thank you, Ayesha. Thank you. Who would like to build on from what Ayesha’s just said? OK, Tahira.
Tahira: We agree with Ayesha. The main priority was that it was right for them to shoot. If they hadn’t shot the dogs, the dogs will die of starvation, and that’s more painful than dying of a shot, because, if you die of a shot, it’s only painful for like one second.
Teacher: OK. OK. So Tahira’s table have just said it was better that the dogs were shot rather than left there to die of starvation, because that would have also been a long, painful death. I don’t know. Malala, do you want to add to that?
Malala: So, on my table, everyone, except for me, agreed on killing the dogs, and Moien said that the dogs would be useful for meat and clothing. The clothing would be able to be used for warmth, and that’s it. And the heat would give the crew members more energy and protein, and it would be healthy for them rather than starving.
Conclusion
I intimated at the outset that I should be using the dataset to argue strongly for the educational significance of classroom group work, and I want now to rest my case. Based on the results I have reported, it is clear that group work amongst students can make a positive contribution to student attainment, both with general reasoning and with all three subject-specific and immensely significant SATs. Moreover, the features that are needed to ensure this contribution are not inaccessible ideals; shortly before the current, pandemic-related closure, they were already widely present in classrooms. Indeed, when the minimum score on the Group Quality Index was 1.55 and only 1.00 would signify total absence, the implication from the data is that all classrooms were already using the features to some extent. When the maximum score was 2.74 out of a possible 3.00, it looks as if some classrooms were already exemplary. In any event, high quality group work is self-evidently manageable within the current education system, and the dividends once high quality is achieved are clear.
In addition, group work is a pre-requisite for beyond-group sharing and the results have revealed such sharing to have two-fold significance. Firstly, it has benefits for student attainment in its own right: the more it occurred, the better the students performed on a science test, a reasoning test, and two of the three SATs. Secondly, when it involves whole-class plenaries where most (ideally all) groups report back and see their reports as contributing to a bigger picture, sharing facilitates dialogue with teachers where students are highly participative, and which is rich in elaboration and querying. Given that these are precisely the features which analyses of the dataset have pinpointed as productive when teachers are involved, the implication is that via beyond-group sharing group work can trigger beneficial interaction with teachers.
Indeed, the implication is that via beyond-group sharing, an integrated approach can be taken to optimising group work amongst students and optimising interaction with teachers, and this surely is important. The research informs and was informed by the literature on dialogic pedagogy, and that literature is replete with failed attempts at classroom implementation due to the challenges involved (e.g., Hennessy & Davies, 2019).
I have already pointed out that high quality group work is demonstrably manageable in classrooms. I now wish to propose that because beyond-group sharing allows group work to be integrated with high quality teacher-student dialogue, a route to making the latter equally manageable also becomes clear.
There is, in short, a potentially seamless and straightforward pathway from group work through beyond-group sharing to elaborative and querying dialogue with teachers, with every reason to anticipate all three providing independent and added value for student attainment. With this recognised, it now seems beyond doubt that group work can be a major positive contributor to classroom practice, and to re-iterate that we should never willingly say goodbye.
References
Galton, M., Hargreaves, L., Comber, C., Wall, D., & Pell, A. (1999). Inside the primary classroom: 20 years on. London: Routledge.
Hallahan, G. (2018). Why it’s time to say goodbye to group work. Times Education Supplement, March 10th.
Hennessy, S., & Davies, M. (2019). Teacher professional development to support classroom dialogue: Challenges and promises. In N. Mercer, R. Wegerif, & L. Major (Eds.). The Routledge international handbook of research on dialogic education (pp. 238-253). London: Routledge.
Howe, C., & Mercer, N. (2007). Children’s social development, peer interaction and classroom learning. The Primary Review (Research Survey 2/1b). Cambridge: University of Cambridge.
Howe, C., Tolmie, A., Thurston, A., Topping, K., Christie, D., Livingston, K., … Donaldson, C. (2007). Group work in elementary science: Towards organisational principles for supporting pupil learning. Learning and Instruction, 17, 549-563. https://doi.org/10.1016/j.learninstruc.2007.09004
Howe, C., Hennessy, S., Mercer, N., Vrikki, M., & Wheatley, L. (2019). Teacher-student dialogue during classroom teaching: Does it really impact upon student outcomes? Journal of the Learning Sciences, 28, 4-5, 462-512. https://doi.org/10.1080/10508406.2019.1573730
Howe, C. (2020). Strategies for supporting the transition from small-group activity to student learning: A possible role for beyond-group sharing. Learning, Culture and Social Interaction. Advance online publication https://doi.org/10.1016/j.lcsi.2020.100471
Kutnick, P., & Blatchford, P. (2014). Effective group work in primary school classrooms: The SPRinG approach. Dordrecht: Springer.
Vygotsky, L. (1998). The collected works of L. S. Vygotsky. Volume 5: Child psychology. New York: Plenum Press.
Image: Pixabay