How you are assessed.
Assessment at AIPC is written before the teaching it tests, published before you sit it, and reported by skill rather than as a single number. This page explains each of those three choices, because each of them costs the College something.
Marked separately, and reported separately.
One aggregate mark hides the thing a language learner most needs to know. These four are scored apart and stay apart on the record.
Listening
Understands speech at natural pace, including unfamiliar accents and imperfect conditions.
Reading
Reads for argument and detail, not only for gist, across registers.
Speaking
Speaks with control of grammar, pronunciation and register, in real time.
Writing
Writes to a purpose and an audience, and revises.
A learner who reads well and cannot be understood aloud has a serious, specific problem, and an averaged mark of 74% conceals it perfectly. Separating the skills is what makes the weakness visible early enough to work on.
6 qualities the programme claims to develop.
Skills are what you can do with the language. Competencies are what the College claims about you as a communicator, and each one is mapped to the specific assessments that evidence it.
Clarity
Understood the first time, by the audience actually present.
Command
Controls the language rather than being carried by it.
Judgement
Chooses register, channel and moment; knows what not to say.
Reason
Constructs an argument, tests it, concedes what should be conceded.
Bearing
Holds a room, a call, a difficult conversation.
Reach
Communicates across cultures, and across the distance between expert and layperson.
A competency with no assessment behind it is a marketing adjective. Each of these is tied to named assessments, and the mapping is inspectable rather than asserted.
Four kinds of assessment, each doing a different job.
A quiz
Checks retention of what the module taught. Machine-marked, immediate, and low-stakes by design — its job is to tell you what to go back to, not to judge you.
An assignment
Produces something: a piece of writing, a recording, a task completed. Marked against a published rubric by a person.
Self-checks
Not marked and not recorded against you. They exist so you can find out what you do not know without it costing anything.
Recorded speech
Speaking is assessed by recording you speaking. There is no substitute, and a programme that assessed speaking on paper would not be assessing speaking.
The rules that produce a mark.
Three different rules look like a pass mark and govern different things: completing a module, passing a level examination, and earning an honour. All three are in force. The table says which does what.
| Rule | Status | What it governs |
|---|---|---|
| 70% to complete a module | In force | The threshold the platform applies today when marking a module complete. It is set in the College’s configuration, not written into the code, so that changing it is a recorded decision rather than a deployment. |
| Level examination: 70% overall, no criterion below 50% | In force | The summative standard. A single aggregate would let a learner pass while failing outright on one dimension; the rubrics already score criteria separately, so the floor costs nothing to enforce. |
| Pass at 70%, no skill below 60% | In force | The award standard, adopted 14 August 2026. It has been applied to nobody, because nobody has yet been assessed. |
| Merit at 80%, no skill below 70% | In force | As above. |
| Distinction at 88%, no skill below 80% | In force | As above. |
| High Distinction at 94%, no skill below 88% | In force | As above. |
Most frameworks let a strong skill compensate for a weak one without limit. This one does not: an overall mark cannot carry a skill below its floor. The reason is plain — a graduate who writes excellently and cannot be understood aloud has not mastered English, and a certificate saying otherwise is one the College would have to defend the first time an employer met them. The floors cost nothing to enforce, because the skills are already marked separately.
You see the criteria before you are marked by them.
Before the assessment, not after
Every assignment carries its rubric with it. Being told afterwards what you were being judged on is not feedback; it is an explanation of a verdict.
The same shape everywhere
All sixty assignment rubrics follow a single published policy on criteria, weighting and band descriptors, and an automated check fails the build if one drifts from it. Consistency across a programme is not achievable by good intentions.
No automated grading of written or spoken work
The College has no automated grading engine and does not claim one. Quizzes are machine-marked; everything that requires judgement is judged by someone.