There is always a start – the start of a day, the start of an important meeting, the start of a relationship. The way you start something matters, as it often creates the attitude and influences that come next. A ‘good’ start can put you in the right mind frame and make the experience productive. A ‘bad’ start can make you feel like you are hitting every possible roadblock…

When it comes to groupwork, everything starts with forming the group.

How do you usually go about group formation? Do you use random allocation or allow students to pick their own groups? Or do you use a different method?

Most instructors seem to be unsure what the best way to allocate groups is. So, my dear reader, sit back, grab a drink and let me entertain you with a literature overview of groupwork allocation. A spoiler: there are no clear answers.

Chapter 1: Let’s all use random allocation!

Once upon a time some scholarly commentators (e.g. Sharan and Sharan, 1992) became concerned about a common practice of students forming their own groups. They felt that the high-achieving students were choosing other high-achieving students, and that weak students were ‘left behind’. In other words, the rich were getting richer, while the poor were getting poorer.

> Who wouldn’t want to create ‘fair’ conditions?So, with the most noble intentions at heart, they declared that students should not be allowed to form their own groups, and that random allocation was, in fact, the fairest way to form groups. This made ‘random’ allocation pretty popular and led many learning and teaching consultants (myself included) to promote it among teaching staff. How could you not? Who wouldn’t want to create ‘fair’ conditions?

Chapter 2: Never use random allocation!

> randomly allocated groups do not outperform other groupsThe problem was however that many students simply hated random allocation and felt at the mercy of a Russian roulette. A relatively recent (Mantzioris & Kehrwald, 2014) meta-analysis of 11 different studies comparing different group allocation methods suggests that random allocation is… the worst thing you can do. Not only do students dislike random selection, but, importantly, randomly allocated groups do not outperform other groups. Mantzioris and Kehrwald’s (2014) conclusion is: stay away from random allocation.

Chapter 3: Shall we play match-maker?

> They call on academics to take a match-making role and come up with an ‘algorithm’ for choosing a groupMeanwhile, back in academic journals…. Several researchers were advocating ‘engineering’ good groups by considering students’ learning styles (e.g. Honey and Mumford Test), students’ complementary skills, personal characteristics, previous performance, gender, culture, etc.  (see, for instance, Huxham, M & Land, R, 2000). They call on academics to take a match-making role and come up with an ‘algorithm’ for choosing a group. What can be easier? As if groups are not ‘complex systems’ with gazillion possible problems, and we (teaching staff) can just set 1-2 parameters, and ‘voila’!

Indeed, the results of the ‘comparative’ studies don’t seem to show significant differences between randomly allocated and ‘engineered’ groups. Students were either ‘slightly’ more satisfied when they were allocated in groups based on their grades (Muller, 1989) or didn’t feel much different (see Huxham and Land, 2001). As for the Muller (1989) study, wasn’t allocating students in groups based on their grades achieving the very inequality that the profession wanted to avoid? The rich getting richer?

Chapter 4: A complete U-Turn  

The same meta analysis (Mantzioris & Kehrwald, 2014) suggests that ‘self-selected’ groups outperform the ‘algorithm’-selected groups. So there you go. A complete U-turn.

It might be because students have their own reasons for choosing group partners, which can be more effective than the ‘algorithm’ logic or a teacher’s hunch. Or it might be because having made their own choice, students feel more ownership, and are therefore more invested, in the group if they have actively participated in selecting/shaping it.

So, my dear confused colleagues. There seems to be no clear answer in the literature, and, as every academic papers says, ‘more research is required’. So watch this space.

Chapter 5: What shall we do?

Probably ask yourself

what it is that you are trying to achieve when you get students to work in groups?

Are you after a great ‘product’ (e.g. a group presentation) or are you interested in something else?

Shouldn’t the real reason to use groupwork be to help students develop their ‘working together’ skills and experiencing different group roles? Become aware of group dynamics, face various group problems and think of ways to solve them? In other words, shouldn’t groupwork be about learning to perform in a group and as a group rather than the final product (e.g. a group presentation)?

> try out each of the different team roles, the leader, executor, administrator, etcIf so, then instead of focusing on a ‘good match’ when engineering groups, perhaps we should try to ‘engineer’ different scenarios. For example, wouldn’t it be good if a student got the chance to try out each of the different team roles, the leader, executor, administrator, etc. at least once throughout their studies? Or get students to reflect about a group conflict or their own tendency to behave in a group?  

This is where program-based design (planning your unit in collaboration with other colleagues) can play a major role. Here you say, “let’s agree, colleagues of mine, which units/assessments will develop students’ ability to run effective meetings, manage time and compromise, improve students’ self-awareness, assertiveness, negotiation skills, etc”.

> make your intention clear to studentsIt would also be very good to make your intention clear to students. “I am interested in seeing how you learn to work together in this assessment task…”, you can then give the instructions such as  “for this group task, pick your friends and note how ‘working’ together changes your relationship”. Or, “pick a person you have never talked to before and note how you present yourself to ‘new’ people”. “What role are you trying to perform? Why?” or “pick someone who has a higher/lower grade than you and observe how you and that person behave… “.

There are many possibilities to make groupwork more meaningful if you are prepared to go beyond the content of your individual subject. While potentially uncomfortable, (e.g. I’m a biology teacher, not a ‘groupwork expert’), it is a very worthwhile exercise to become familiar with, as it’s really the only way to develop transferable skills of working together, that, arguably, can make or break students’ future professional lives.

How do YOU usually go about running groupwork? Any thoughts or tips for Teche readers?

Huxham, M & Land, R (2000) Assigning Students in Group Work Projects. Can We Do Better than Random?, Innovations in Education and Training International, 37:1, 17-22, DOI: 10.1080/135580000362043

Mantzioris, E & Kehrwald, B (2014) Allocation of Tertiary Students for Group Work: Methods and Consequences , Ergo, 3 (2)

Muller, T. E. (1989). Assigning students to groups for class projects: An exploratory test of two methods. Decision Sciences, 20(3), 623-634.

Sharan, Y. and Sharan, S. (1992), Expanding Cooperative Learning through Group Investigation, Teachers College Press: Columbia University, New York, NY.

Posted by Olga Kozar

I'm a 'long-term' Mq girl. I did my PhD here and taught on different courses, ranging from 1st year to PhD students. I now work in Learning and Teaching, which I love. I have 2 young kids and a dog, and I love meeting other Mq people, so give me a shout if you'd like to talk 'learning and teaching' or would like to brainstorm together.


  1. […] What is the best way to allocate groups (TECHE) […]


Leave a reply

Your email address will not be published. Required fields are marked *