Where COIL partnerships tend to come unstuck

COIL partnerships rarely fall apart because the faculty stopped caring. They fall apart structurally — through a decision that was never quite made, surfacing as a problem weeks later when there is no time to fix it.

Four patterns recur often enough to be worth naming. This is a general guide rather than an account of any particular partnership.

1. Two enthusiastic people, and no owner

The founding condition of most partnerships is mutual enthusiasm between two academics who like each other’s thinking. That is a genuinely good start, and it is also why nobody owns the dull parts.

The dull parts are: chasing a deadline the other side has drifted past, telling a partner that four of their students have dropped the module, rebooking a session when a strike or a public holiday lands on it, and asking a colleague twice for something they said they would send. Between peers of equal standing, in different institutions, with no reporting line, these things are socially awkward — so they get left.

Enthusiasm distributes credit easily and responsibility badly.

What helps: name one coordinator on each side. Not a leader, and not a hierarchy — a named person whose job includes the awkward messages, so that sending one is a role being performed rather than a friend being difficult.

2. Assessment weighted so unevenly that one cohort disengages

This is the most reliable predictor of a collaboration going quiet, and it is almost always discovered too late to change, because grade weightings are usually fixed in approved module documents months in advance.

When a collaboration is worth thirty per cent of one cohort’s grade and five per cent of the other’s, the second group is behaving entirely rationally when it deprioritises the work. The first group does not experience it as rational. They experience it as being let down by partners who do not care — and that reading, once it takes hold, is very hard to shift.

What helps: compare weightings during planning, not delivery, and get them within sight of each other. Exact parity is rarely achievable across two validation regimes. An order of magnitude apart is close to fatal. Where the gap cannot be closed, say so openly to both groups at the start, so the lighter commitment is a known condition rather than a perceived slight.

3. No plan for the students who go quiet

In any cohort some students disengage — illness, paid work, caring responsibilities, anxiety about writing in a second language, or simply losing the thread. This is ordinary. What makes it harder in a cross-institutional group is that the usual safety net is missing.

In a normal seminar group a tutor notices a quiet student within a week or two. In a COIL group each tutor sees only their own half, and the students carrying the extra load are usually reluctant to report a peer they have only met on video — particularly across a language or cultural difference where they are unsure whether silence means disengagement or something else entirely.

What helps: three things, all cheap. A shared workspace both tutors can actually see into, so absence is visible rather than reported. A scheduled check-in at the midpoint where each group says how it is going, which makes raising a problem routine rather than an accusation. And an agreed rule, given to students in advance, for what a group does when a member goes dark — so the remaining students know they will not be penalised for someone else’s absence.

4. Nothing written down, so the partnership leaves when a person does

A partnership that works for two or three years accumulates a great deal of knowledge: which week the Norwegian reading break falls, that the shared deliverable needs a hard deadline before one institution’s exam board, that groups of four work and groups of six do not, and the name of the person in IT who can actually issue guest accounts.

Almost none of this is recorded. It lives in two people’s memories and two email archives.

When one of those two people takes a headship, moves institution or goes on leave, the partnership usually does not get handed over. It quietly lapses, and a few years later someone proposes building an international collaboration with a university they do not realise their own department already worked with.

What helps: a two-page record, updated once a year, held somewhere both institutions can reach. The shared outcome, the dates that worked, the assessment agreement, the tools, named contacts on both sides including administrative ones, and a short note on what changed this year and why. It is an hour of work annually and it is the difference between a partnership and a friendship that happened to produce a course.