The principle: marks attach to artefacts, not to impressions¶
Every mark in this rubric must be justifiable by pointing at something in the submission — a section of
an artefact, a table, a record. Where a descriptor cannot be evidenced from the submission, it cannot be
marked, which is why this rubric deliberately excludes effort, ambition, and polish as separate dimensions.
The consequence is worth stating plainly: a modest project with a complete, consistent, honest record
scores higher than an impressive one whose record is thin. That is not a concession to weaker projects; it
is what the course is for.
Change request, decisions, status reports, the response to what went wrong
6
Risk and trade-off analysis
15
Risk register, response plan, schedule and cost analysis, scope decisions
7
Professional documentation
5
Clarity, structure, sourcing, labelling of invented data
8
Reflection and lessons learned
10
Final report, lessons register, retrospective, self-assessment
Weighting rationale. Consistency carries the highest weight because integration is what the capstone
exists to assess: any single artefact can be produced in isolation, but only a managed project keeps eight
documents agreeing with each other. Documentation carries the lowest because it is a reporting skill, not
the subject — professional adequacy is enough, and polishing beyond it earns nothing.
Terms used loosely or incorrectly; processes described but not applied; no distinction between related concepts (constraint vs assumption, risk vs issue)
2 · Developing
Correct use of most terminology; techniques applied mechanically, with the purpose stated only in general terms
3 · Competent
Terminology used accurately throughout; each artefact's purpose is evident from what it contains; deliberate choices between approaches are explained
4 · Strong
As Competent, and the student reasons about when the technique would not apply, or where its limits lie, with reference to their own project
Artefacts contradict each other on facts: dates, scope, effort, or stakeholders appear differently in different documents
2 · Developing
Broad consistency, with unexplained discrepancies — a deferred item still in the WBS, an estimate that does not reconcile with the budget
3 · Competent
Artefacts agree on every material fact; where they differ, the difference is explained
4 · Strong
As Competent, and the submission shows the reconciliation happening — a dated update, a decision log entry, a re-baseline record — so the reader can see how consistency was maintained rather than assumed
Note: level 4 requires evidence of the process, not only a consistent end state. A submission that was
consistent only because it was written in one sitting, at the end, will not have that evidence — and
nothing in the rubric asks the student to disclose which of those happened, so the marking rests on what the
artefacts themselves show (dated records, decision entries, version history).
Assumptions absent, or stated as facts; no consequences recorded if they prove false
2 · Developing
Assumptions listed separately from constraints, with consequences for some; none tested or owned
3 · Competent
Assumptions recorded with their consequences, and the significant ones tested, evidenced, or flagged for confirmation — with the outcome recorded
4 · Strong
As Competent, and at least one assumption was found to be wrong and the response is visible: the artefact updated, the impact assessed, and the lesson routed
Note: the levels do not require an assumption to be wrong. They require that where one was, the response is
visible — and where none was, that the testing is evidenced.
Risks listed as categories with no owners or triggers; no trade-offs acknowledged
2 · Developing
Risks in cause–event–effect form with owners; responses present but not costed; trade-offs implied rather than stated
3 · Competent
Risks scored consistently with triggers and residual exposure; at least one costed trade-off (scope, schedule, quality, or risk) with the consequence accepted by a named person
4 · Strong
As Competent, and the analysis identifies a secondary risk created by a response, or a shared cause behind several risks, and quantifies its effect
Documents that a third party cannot follow; unlabelled invented figures presented as fact; no sources for external claims
2 · Developing
Readable, with inconsistencies in structure; some figures unlabelled or unsourced
3 · Competent
Clear, structured, and navigable; invented figures labelled as the author's own assumptions; external references verifiable
4 · Strong
As Competent, and the documents are usable by a successor — an index or a note explains what each artefact is and what to read first
Note: level 4 is about usability, not aesthetics. No mark in this rubric depends on formatting choices,
templates, or presentation beyond professional adequacy.
Dimension 8 — Reflection and lessons learned (10)¶
Level
Descriptor
1 · Not yet
Lessons are generalities ("communicate better"); no reference to what happened in this project
2 · Developing
Lessons reference real events but name no change to practice, or name changes with no owner
3 · Competent
At least four lessons, each with the situation, the insight, an owner, and a change to practice; the final report explains variance rather than excusing it
4 · Strong
As Competent, and lessons are routed to where a future project will meet them (a checklist, a template, an estimating basis), with an honest disposition — including "not adopted" with a reason — and the self-assessment names a genuine weakest dimension
Competent throughout with several Strong dimensions, including consistency and reflection
Merit
70–84
Competent throughout, with no dimension below Developing
Pass
50–69
Most dimensions at Developing or Competent, with at least one dimension lower
Not yet
below 50
Multiple dimensions at Not yet, typically because artefacts are absent or inconsistent
Where a submission is borderline, the tie-breaker is the consistency dimension: which of the two versions
of an internal contradiction was written first is not knowable, but a project whose artefacts agree is the
one a successor could actually use.
Cite the artefact and section for every dimension you mark above or below Competent.
Mark the reasoning, not the conclusion. A well-argued decision that differs from your own preference
scores on the same scale as one that matches it.
Do not deduct for a project that went wrong, provided the record shows it was noticed and answered. A
capstone with no recorded variance is more likely to be under-reported than perfect, and the rubric's
dimension 3 rewards the dated records that make such a project's history legible.
Treat the self-assessment as evidence. An over-generous self-assessment without citations is itself a
finding, and it belongs in the feedback rather than in the total.
Use the same descriptors for self-marking and instructor marking, so students can predict their
result — the rubric's value depends on it being the same instrument in both hands.