Group projects, and who ends up carrying them

Nobody plans to do four people's work at two in the morning, and nobody plans to spend six weeks contributing almost nothing to something with their name on it. Both happen constantly, in the same groups. The usual explanation — some students are conscientious, some are lazy — describes the outcome, not the cause.

The person who takes over

There is a kind of waiting that is genuinely intolerable. The deadline is visible, the work is not moving, three people have not replied, and the discomfort of that gap is worse than the cost of doing it yourself. So they do it. Often well, often fast, often at midnight.

From inside, this is not seizing control. It is being the only person who can see the cliff. A real risk is being avoided too: a shared mark means someone else's silence becomes your grade.

The price is paid quietly. They finish exhausted, resentful, and certain that group work taxes the people who care. And they have, without meaning to, removed the thing the others were about to start.

The person who goes quiet

This looks like disengagement and often is not. The first meeting produced no clear division of work, so it was unclear what to begin. By the time it was clear, someone had begun it. Asking what should I do three weeks in is an admission, and the longer the silence runs the harder it is to break.

Some of this is timing rather than motivation. If you need to see the shape of a thing before starting, a fast group makes three decisions before you have finished forming a view, and a better contribution arriving afterwards is worse than useless. Some of it is that the quickest person took the visible parts, leaving the residue.

The price here is also quiet: a mark for work they did not do, nothing learnt, and more evidence for a story about not pulling their weight. Each position produces the other.

What to agree in the first meeting

This is fifteen minutes and it changes most of what follows.

• Split by deliverable, not by topic. Not you look at ethics but 600 words on ethics, in the shared document, by Tuesday the 14th. An artefact, an owner, a date.
• Set internal deadlines two weeks before the real one. The submission deadline is the point past which nothing can be corrected, not a working date.
• Agree where the work lives. One shared document, edit history on. That does more for visibility than good intentions ever will.
• Agree what happens if a piece does not arrive. Say it while everyone is comfortable: we message the person once, and if it is still missing at the next meeting we reallocate it. Agreed in advance, the awkward conversation is a procedure, not an accusation.
• Write it down and send it round. Ten lines. A group with this in writing behaves differently from one that discussed it.

When it has already gone wrong at week three

If you are carrying it: stop adding work for forty-eight hours and send one message naming specific unclaimed pieces with dates, rather than asking whether anyone can help. General appeals get general agreement and no work. Then email the module leader — not to report anyone, but to state what has been produced and by whom, and ask what the process is. At week three that is information. At week nine it is a complaint.

If you have gone quiet: pick the smallest unfinished piece, do it, and send it with no apology and no explanation. Here is the methods section. Re-entry is easier through delivered work than through a conversation about not having delivered.

What the setting can do

Most of this is settled in the assessment design, where departments have a choice.

Individual accountability within group work is a known design feature: an individually marked component alongside the group mark, a required contribution statement, or assessment of a defined section by its named author. Where it exists, both positions ease, because the shared mark stops being the only signal.

The other substantive ask is that groups be given a structure rather than told to organise themselves — allocated roles, a work plan due in week two, a midpoint check with a tutor. Left to invent it, most groups invent it badly, and the students who pay are the same ones each time. Both are reasonable to raise in module feedback or through a course rep, and put as design questions rather than complaints they are far more likely to land.

Small things to try this week

1. Post the deliverable list with names and dates in your current group, even if nobody asked.
2. Move the work into one shared document with visible edit history.
3. Put an internal deadline in the calendar two weeks early, and tell the group that is your date.

If the same pattern repeats across several groups with different people in them, that is useful information, and it points at how the assessment is built rather than at your classmates.