Why (the hell) should students work in groups?

//Why (the hell) should students work in groups?

In a recent TES article James Mannion and Neil Mercer make the following claim:

In not using group work, students are denied the chance to develop skills that can not only help them perform better in schools, but which are also vital for their future employment prospects – not to mention the realisation of a more fully participatory democracy.

First, let me state that I see nothing inherently wrong with pupils working collaboratively. Like any method of working, it has its time and place. But I was led to believe that unless a lesson contained an element of groupwork it could not be a good lesson. It’s not enough to claim that ‘no one really believes this’. I did. I come a from a generation of teachers raised to believe groupwork is inherently good and for that reason alone it deserves to be questioned, prodded and poked.

It’s easy to poke fun at groupwork (just google “groupwork+funny”) but let’s consider what the problems might actually be.

Way back in 1913 French agricultural engineer, Maximilien Ringelmann discovered something rather surprising: the productivity of a group decreases as its size increases. If we pull a rope all by ourselves we tend to pull our hardest but as soon as others pop up to help us out we slacken off. Even though they are likely to be unaware of this slackening of effort, everyone pulls a bit less hard. The bigger the group, the greater the tendency for social loafing. The Ringelmann Effect suggests that when we’re part of a group we believe every other member is doing the hard work. We can take it easy because our lack of effort won’t be exposed. Unconsciously, we rely on those around us to pull out the stops to get the job done. This phenomenon is well known to teachers. It ought to be reasonable to expect a group of four pupils to produce four times as much work collectively as they would produce alone. In fact they tend to produce less together than they might alone.

Not only that, working in groups actually makes us less creative. In particular, ‘brain-storming’ (or ‘thought-showering’ if you prefer the more politically correct synonym) actually limits our capacity to come up with interesting ideas.

So why do we bother? Why have pupils work together at all?

Let’s have a think about Mannion & Mercer’s claims. Firstly the idea that group work develops skills which lead to ‘better performance’ in schools. I’ve heard this stated a number of times and usually the evidence I’m directed to points at Robert Slavin. In Cooperative Learning: What makes Groupwork Work? (Which Mannion & Mercer cite) Slavin runs through the evidence supporting many and various different forms of group (or collaborative learning) but the gains seem slight and the studies wooly. In terms of ease and effectiveness it seems preferable not to bother so there would have to be some pressing reason to make teaching more difficult than it needs to be.

So how about future employment prospects? Will group work give students a much-needed edge in the world of work? The argument here is that “by learning how to work well in teams, young people are acquiring the soft skills that are valued by employers throughout the world, and promoted by schools in high-achieving countries such as Singapore.” I understand the reasoning but it rests, I think, on the same fallacy that suggests independent learning will result in independence. In order to perform like experts, we must first be experts. Emerson tells us that “Every artist was first an amateur”. Rushing students into situations where they are expected to behave like experts misses the fact that they don’t yet know enough to do so. That doesn’t mean that there’s no point in students working together, but it does undermine the idea that groupwork creates better workers.

Finally, there is the thought that groupwork gives students an insight into democratic participation. Leaving aside the idea that we could simply teach students about democracy if we felt it was important for them to know about, I’m dubious about this proposition also. Enormous amounts of effort and expertise must be applied to get groups working in a functional way. Mannion and Mercer (M&M from now on) themselves identify that the problem with groupwork is that teachers don’t feel up to it. They say that “teachers and students just haven’t been taught well enough how to make it work.” This may be so, but is it really beneficial to expend the resources on such training? What of the opportunity cost? And unless such a commitment is made, students experience of groupwork will not be an oasis of democratic peace and love. Bad groupwork is worse than almost any other classroom sin, ending in the tyranny of the strong and the persecution of the weak.

One of M&M’s arguments for groupwork is that it develops students’ oracy. I’m a big fan of developing oracy, although probably for different reasons – I consider learning to speak the language of academic success a necessity if you are to be academically successful. Students get plenty of practice talking already – what I want is for them to have structured opportunities to take part in discussion, modelled and scaffolded by the teacher. This is least likely to happen in small groups and mostly likely to take place in either paired discussions or whole-class debates.

So is there a point to groupwork?

We’ve all encountered pupils who struggle to answer questions and come up with ideas. Left to their own devices they sit, head on desk and dream of being somewhere better. We know that simply getting them to discuss some possibilities with the student sitting next to them can be sufficient to jolly them along. Maybe they haven’t become more creative – maybe this just gives them less of an excuse for doing nothing? Who cares: it gets them working. I’m sure we can all cite thousands of examples from our own lives of occasions when a simple conversation with a friend or colleague opened up new possibilities or pointed us in previously unexplored directions. The point of collaboration is that it opens us to the ideas of others. But then, so does reading books.

2018-09-23T15:17:12+00:00February 11th, 2015|Featured|


  1. […] Read more on The Learning Spy… […]

  2. nphillipson February 11, 2015 at 6:51 pm - Reply

    I think dialogue supports learning in a variety of ways. It helps us to make meaning from factual information, it allows misconceptions to be surfaced and challenged and perhaps most importantly it helps us to constructively and critically engage with other perspectives, seeing them as a ‘call to thinking’ – an opportunity to learn something new, or to test the validity of your existing view (I have, perhaps a little clumsily, expanded on these thoughts at http://21stcenturylearners.org.uk/?p=556). Dialogue can take place in both whole class and small-group formats and both have merits. Guidelines such as those suggested by Mannion and Mercer,and the ‘4Cs’ framework at the heart of P4C, can be a foundation for effective dialogue, and I would argue that the investment in time needed to embed them is well worth it.

    • David Didau February 11, 2015 at 6:55 pm - Reply

      I’m a huge fan of dialogue, but I think groupwork is more likely to impede it than anything else.

      • nphillipson February 11, 2015 at 7:06 pm - Reply

        I guess it depends how we are defining group work. I think properly structured tasks that require meaningful dialogue in small groups, underpinned by the guidelines mentioned above, can be very powerful (teaching for effective dialogue as well as through it is an important idea too). I entirely agree that unstructured and unpurposeful (is that a word?) group work can, and usually does, lead to a significant percentage of the class learning very little!

        • David Didau February 11, 2015 at 7:49 pm - Reply

          Small groups are fine if they are groups of two. Any larger tends to waste time in some way (whether it’s teacher’s or students’).

  3. Debbie Evans (@Apidae77) February 11, 2015 at 7:19 pm - Reply

    ‘group work’ in the traditional sense is exactly as you describe. However, I have recently found that properly structured paired and team learning tasks to be extremely effective teaching methods.

    • chrismwparsons February 11, 2015 at 9:18 pm - Reply

      Certainly I would greatly advocate the strategic use of paired talk. In my experience, larger group work is of most use for training pupils to do larger group work…(circularity of comment deliberate!) You can condition pupils to perform quite efficiently as part of structured teams, but I’m not sure that for the vast majority of occasions they really would learn more by operating like this.

      • David Didau February 11, 2015 at 10:04 pm - Reply

        Yes, preceisely that. Nicely put.

      • Doreen March 10, 2015 at 9:34 pm - Reply

        I like pairs to answer focus questions then to compare with another pair. Enables them to look for similarities and differences in ideas and often to double the number of points they can feedback.

  4. Dominic Salles February 11, 2015 at 11:08 pm - Reply

    Completely agree. Really interesting that Teach Like a Champion, by Doug Lemov, has no section on group work. His value added teachers seem not to do it.

  5. jfin107 February 12, 2015 at 9:12 am - Reply

    In music education pupils often learn in groups because playing in groups is a traditional form of music making. Providing pupils with ‘talking points’ in order to support the process stimulating cumulative talk and the working out of musical ideas is valuable. There are some examples in ‘Interthinking’ (Littlejohn and Mercer 2013).

    • David Didau February 12, 2015 at 8:03 pm - Reply

      Correct me if I’m wrong but surely learning to play a musical instrument to any standard requires hours of individual practice? Any ensemble practice can only come after the individual is competent, no?

      • jfin107 February 12, 2015 at 8:36 pm - Reply

        In school it is common for instrumental learning to be a group arrangement. This is directed by a teacher. Children will also engage in individual practice as you point out. Also common in music educational practice is group imrpovisational activity which may result in the forming of a composition. This is hopefully scaffolded through whole class instruction. In the group musical ideas are developed, evaluated and refined. Here is musical dialogue counterpointed with talk. With the help of ground rules and appreciation of different types of talk pupils learn the nuts and bolts of music, learning to think in sound and to think about it. Much depends upon musical genre, style, tradition. If the tradition is Samba, for example, a group of 7 or 8 can successfully develop a Samba performance. Developed from the Littlejohn and Mercer research talking points can be used to challenge the thinking process and raise levels of meta-cognition.
        Thank you for your interest.

        • jfin107 February 13, 2015 at 6:09 am - Reply

          I recently observed a skilful RS teacher deploying group work who had taught his pupils to distinguish between Mercer’s three types of talk and with signals lead pupils to switch between these. All this supported the pupils thinking and their writing.
          There is much here for teachers to investigate, and as some have found, a valuable area for their action research projects.
          Now you have got me thinking about: thinking in sound (music); thinking with music; thinking about music and the role of language in this powerful ensemble. Thank you.

  6. chrismwparsons February 13, 2015 at 7:35 am - Reply

    Ok, How about this as a heuristic for deciding whether to use groups?:

    Is it something that can’t be done as well by pupils working individually?
    Is it something which requires as many pupils as possible to actively practice something, so it won’t be as useful if done as a whole class group?
    Does it need any more than a single partner to enable all elements to work?

    It seems clear that in some circumstances, such as a science experiment, or jfin107’s music groups, or a drama session, you really need people playing different roles to make the whole work, and group-work is an ideal model which will require careful setting-up and training.

    The example above of the RS teacher using group work to get pupils to model the different types of talk to each other is an interesting one. On the surface, you could say that using small groups for this allows more pupils to have actual practice of modelling the kinds of talk. HOWEVER, what also then starts to slip is the element of quality control which the teacher would have if this was done as a whole class activity. The pupils in the groups guessing the talk-type will only know if they are right when the person modelling the talk tells them; but what if the person modelling the talk actually models it badly or incorrectly? How does the teacher ensure that what is learned is straightened out?

    As is the focus of the Lemov et al book ‘Practice Perfect’, practice, on its own, simply makes permanent, not perfect. If we want pupils to not ingrain poor quality learning, we need as much opportunity as possible for teachers to intervene into pupils practice, correct it and polish it. They can of course do this with whole class interactive teaching, they can do this through marking with individual work, and they can actually do this pretty easily with partner work, as it takes relatively little time to start and stop a useful injection of this, before calling out a few examples to hold up to scrutiny.

  7. jfin107 February 13, 2015 at 11:56 am - Reply

    Perhaps much depends on the extent to which the teacher believes talk to be a form of thinking through which understanding is achieved. Well ordered group work where pupils have learnt the rules and appreciate the benefits of developing thought in this way gives greater opportunity to do thinking through talk that interactive whole class teaching, vital though this is.

    • David Didau February 13, 2015 at 12:53 pm - Reply

      I take your point but I’d argue pairs talk more than larger groups and are easier to manage.

  8. jfin107 February 13, 2015 at 1:29 pm - Reply

    Great opening here for teachers to research the question as it applies to their own practice.

  9. Robert Craigen February 13, 2015 at 5:26 pm - Reply

    Democracy is not the blending of individuals into a collective. It is cooperation of different, and often disagreeing, individuals for the accomplishment of shared goals. The purpose of groupwork, as stated in most educational materials I’ve seen, is the former, and more closely conforms to the ideals of socialism and collectivism than to that of democracy. Groupwork needn’t involve the blurring of individual work and accountability into the group, but it takes much effort and planning to avoid this, and I don’t see many groupwork advocates who acknowledge or show much sympathy for this concern. I believe this “meme” (i.e., that groupwork is more democratic than individual work) is promoted by those ideologically predisposed to favour socialism and who would like to see it conflated with participatory democracy as the latter is universally regarded in a positive light; but they are not the same thing. Yes, socialist societies can be democratic, there is no intrinsic contradiction. But it’s easy enough to produce examples of the other kind — indeed, they seem considerably more common, particularly at the far end of the socialist spectrum.

    While your essay gets into some of the philosophical and practical problems with groupwork, it misses what I consider the biggest problem, it’s effect on students of particular types.

    In particular, groupwork is generally hated by

    (a) The strongest students who resent others freeloading on their efforts and ability and losing the individual recognition they feel they deserve for their accomplishments. Indeed, many strong students are motivated by recognition and the experience of success. To have that blended into the background of the collective takes the steam out of a strong incentive. It may artificially “narrow the achievement gap” but at the expense of those who would otherwise rise to excellence. In my discussions with teachers who are enthusiasts for groupwork, when we get to discussing the effect on the strongest students, I am shocked at how often their response is that this is a GOOD thing: “those kids need to be taken down a notch anyway!” “They’re so cocky — they love to gloat! This’ll make them think of the needs of others!” I wonder if some of them see this as a way of “getting back at” the superstar they had to “put up with” when they themselves were young. What is more shocking, perhaps, is when it is pointed out to them that this is a base motivation for adopting a teaching method, they either don’t get the point, or they get it … and don’t care.

    (b) The weakest students, who feel like a fifth wheel, or in some situations are derided or scorned by students doing the work and who may resent their presence on “the team”. In individual work they must cope only with an adult who monitors their poor performance — an adult whose responsibility is to provide them with expert assistance in overcoming their difficulties. In group work it is exposed for the world to see … and more specifically, for their peers to see. Peers unqualified to provide expert educational remediation. Children can be intentionally or unintentionally cruel, and even when they are not, it can be mortifying to have your academic blemishes exposed in this manner. There is a reason that we generally don’t post individual marks with student names — and that reason is to protect individuals’ privacy who are vulnerable in this way.

    (c) Students with social or communication deficits. This demographic is concentrated in the highest and lowest performing groups, but these kids are found in all ability levels. Such students are uniquely handicapped for this form of learning and it can be excruciating for them to be forced to continually throw the ball with the wrong hand like this, to continually be hampered by their handicap. Tragically, many of these students would otherwise find that they can compensate for their disability by striving for individual excellence in the academic realm. Too much groupwork can take this path away from them.

    Finally, group work is a distraction for all from the subject matter. It forces students, and teachers, to continually negotiate the development and adaptation of social skills with the development of new academic skills. Why, when many students have trouble juggling three balls, must we insist they keep another two or three in the air at the same time?

    There is a time and place for group work, and in moderation it should not overly harm students, and may even add spice to their educational experience. But it is not as “spicy” if it becomes the whole diet. As I like to say, there are main dishes and side dishes in the educational meal, and a good teacher does not confuse the two.

    The conventional classroom does not lack for social interaction. For one, there would always be regular one-on-one intervals with the teacher and/or aides. Further, in the traditional classroom it is common for stronger students — with the right social skills — to be encouraged to help out the weaker ones. I am forever grateful to a teacher who expertly paired me with an academically strong girl in my grade 5 class to overcome a problem I had learning a simple math skill, that put me at the bottom of the class; the timely intervention made a huge difference for me. Today I have a PhD in Math and have taught university math for over 20 years and won awards for my research.

    Finally, the conventional class uses a lot of whole-group instruction and classroom discussions, and sprinkles purely social elements throughout the day, making the classroom into a sort of extended family experience. It is ludicrous to paint such classes as deficient in the social arena.

  10. Robert Craigen February 13, 2015 at 5:41 pm - Reply

    Sorry, after that long comment I realised I had a couple more things to say:

    First, the manner in which a subject is taught should be respectful of the nature of the material itself. As someone mentioned here, science labs are particularly suitable for paired or grouped work. Art or media projects may also be. Less so english composition or math skills. There seems to be an attitude among educationalists that one first should adopt a pedagogical philosophy, then force all instruction to conform to its demands, regardless of the nature of the content in question. I strenuously disagree: pedagogy is downstream of content, not the other way around!

    Second, while there is abundant research showing that learning styles are nonsense (actually I believe what has been shown is that, while students may have what are described as “learning styles”, nobody has yet produced a system of taking advantage of these that performs consistently well in the classroom under test conditions), I am of the view — held, mind you, from anecdotal experience, not because of any research I have read — that there is such a thing as “teaching styles” and further that a teacher teaches best according to their personal teaching style. These are not necessarily fixed; you could develop new teaching styles — but to try to teach according to a style that you are not adept in is to court certain disaster. I would encourage teachers to go with their gut, and what reflective observation of what works and doesn’t work in their class tells them about where their strengths lie. And … if a teacher’s personal “teaching style” happens to involve using a lot of groupwork — then that is probably a good thing for them to do. I would encourage such teachers, however, as with all teachers, to consider the potential downsides of teaching approaches they use, and plan to compensate where necessary.

    I am willing to be proven wrong about teaching styles, but it would take at least 3 well-designed RCTs to shake me from my current perception about the matter.

    • sean February 13, 2015 at 11:28 pm - Reply

      This is a response to group work. I am not as familiar as you with ringelmann, but isn’t there a better example than pulling a rope? Surely a simple task like this where every group member is doing the same thing is’nt comparable to well planned group work in a school learning environment.

      • David Didau February 14, 2015 at 6:47 pm - Reply

        The principle – first observed by Ringelmann – has been termed the Ringelmann Effect because it has since been observed in countless other social arenas, including group tasks in school. My own anecdotal experience is that while it might be reasonable to suggest a group of four should produce work equivalent to four times the output of single individual what normally happens is that they produce a quarter of that an individual would turn out on their own. The psychological effect ‘social loafing’ is well documented and widely agreed to exist.

    • chrismwparsons February 13, 2015 at 11:53 pm - Reply

      Robert…My God – there is so much that I agree with here. Thank you

  11. jfin107 February 14, 2015 at 9:07 pm - Reply

    The recent work of Robert Slavin presents a model of co-operative learning and how through co-operative learning, learning can be enhanced. Slavin discusses four major theoretical perspectives on co-operative learning and achievement and examines the evidence supporting enhanced learning. The Mercer et al perpsective is not one of these. Interthinking (Littejohn and Mercer 2013) works through a social-cultural view of the mind (Vygotskian). Interthinking offers the theory and empirical evidence. It is likely that the theory of mind held by each of us will lead to particular value positions on group work. Resonses above show this well. That there are alternative positions is on balance I think a good thing.
    In Slavin’s recent paper in ‘Education 3-13’ he concludes ‘Co-operative learning has the potential to become a primary format used by teachers to achieve both traditional and innovative goals. Research must continue to provide the practical, theoretical, and intellectual underpinnings to enable educators to achieve this potential.’ (12)
    This statement is welcome.

  12. […] recently posted some thoughts on what group work is and isn’t good for. At no point did I say it was good for nothing (although predictably my opinion was caricatured as […]

  13. jfin107 February 15, 2015 at 11:55 am - Reply

    David’s blog and the ensuing discussion has led me to wonder whether in such discussions some thick descriptions from classroom practice within particular subject domains would help understanding and serve to supplement generalised contributions. The teaching of knowledge, skills and dispositions within particular subject domains carry particular sets of values which are likely to contextualise the value of group work in different ways.

  14. […] Didau (@learningspy) has written a response to this article by Neil Mercer and I, entitled Why (the hell) should students work in groups? He raises a number of points I would like to respond to directly – as such, this post will adopt […]

  15. chris walton March 6, 2015 at 12:26 pm - Reply

    There’s an old Suffolk farming saying (pointed out to me by Michael Marland) that goes: one boy’s a boy, two boys is half a boy, and three boys is no boy at all, which kind of backs up what you’re saying.

  16. K April 20, 2015 at 8:25 pm - Reply

    This is an interesting read, as I myself have considered group work to be beneficial for children; however through implementing this within practice I have seen both negative and positive effects. As you rightly said there are problems with group work which lead to it not adding anything to lessons. Factors such as members of the group not pulling their weight Baines, Blatchford, Kutnick, Chowne, Ota and Berdondini (2009), as well as the need to accommodate different working speeds (Kershner, Warwick, Mercer and Staarman 2014) could be the basis for these difficulties. These are all aspects which have potential to be overcome in order for more effective group work. However I feel these implications are outweighed by the benefits associated, such as the opportunity group work offers for children of all abilities to extend and develop their thinking and reasoning. Baines et al (2009) suggest that ‘cognitive conflict’ produced by conversing with others allows exploration and construction of ideas. Though this is the case, simply putting children into groups does not necessarily make their work better, therefore a collaborative approach (Jolliffe 2007) as oppose to a general group approach could be implemented instead. This would allow children to take on roles, consequently providing a purpose for each child to contribute, based on the knowledge that they need to work collaboratively to complete work (Jolliffe 2007).


    Baines, E, Blatchford, P, Kutnick, P, Chowne, A, Ota, C and Berdondini, L (2009). Promoting Effective Group Work in the Primary Classroom: A handbook for teachers and practitioners. Oxon: Routledge

    Jolliffe, W (2007). Co-operative Learning in the Classroom: Putting it into Practice. London: Paul Chapman Publishing.

    Kershner, A, Warwick, P, Mercer, N and Staarman, J.K (2014) Primary children’s management of themselves and others in collaborative group work: ‘Sometimes it takes patience …’. Education 3-13. 42(2)

  17. […] (2013) and Vai and Sosulski (2016) mention group work as a positive experience. Yet (Didau, 2015) questions the usefulness of group work. Admitedly he is talking about adolescents in a school, who may be less mature in their approach […]

  18. […] paired work during which she can chat away with another like-minded student. There needs to be a much better justification for group work than that a few children like […]

Constructive feedback is always appreciated

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: