Assessment planning resource

Process Evidence Planner

Plan only the evidence needed to support a defined learning outcome in AI-assisted programming.

A planning aid, not a standard. This proposed TraceYield method helps make evidence choices explicit; it is not a validated assessment instrument.

Learning outcome purpose

Task & context opportunity

Selected evidence what adds value

Explain & verify bounded checks

Lecturer judgment human decision

Blank planning template

01—06 · design the evidence set

Learning outcome

What should the student demonstrate? State the assessable knowledge or skill.

Task & context

What work will make the outcome observable? Note constraints and AI-use conditions.

Final artifact

What will the submitted implementation already show against this outcome?

Selected trajectory / process evidence

Which small part of the development process adds information the artifact cannot provide?

Explanation / demonstration

What focused explanation, adaptation or demonstration may be relevant?

Verification

What check is relevant to the outcome, and what is its scope?

Interpret with care

07—08 · lecturer responsibility

Lecturer assessment

How will you interpret the evidence against the stated outcome and rubric?

Uncertainty / follow-up

What remains unclear? Would a proportionate question or transfer task help?

Layer 1Development / process evidence

Layer 2Evidence relevant to a learning outcome

Layer 3Lecturer assessment and judgment

Process Evidence Planner · completed illustration

Equipment-booking service

A worked example using the same planning sequence. Evidence is selected for one specific programming outcome.

Illustrative example

Not student data. Not classroom research. Not evidence from a TraceYield pilot.

Plan the evidence set

01—06 · task to verification

Learning outcome

What should the student demonstrate?

Design and implement a data-integrity rule for a concurrent web application, and justify how the rule is verified.

Task & context

What work and conditions make it observable?

Prevent two users booking the same device during overlapping time periods in an existing service. AI use is permitted under the assignment brief.

Final artifact

What does the submitted implementation show?

A service change, database constraint, conflict handling and integration tests for overlapping bookings.

Selected trajectory / process evidence

What adds information beyond the artifact?

The initial check-then-insert proposal, a concurrent test failure, and the revised hypothesis that the invariant belongs at the database boundary.

Explanation / demonstration

What focused response could be relevant?

Explain the race condition, then adapt the design for a second device type or a changed booking rule.

Verification

What check, with what scope?

A targeted concurrent-request test and booking regression tests; note their environment and coverage limits.

Uncertainty / follow-up

This set alone does not establish general competence, authorship, independent work, intent, misconduct or a grade. If the outcome remains unclear, the lecturer chooses a fair, outcome-aligned follow-up.

Evidence supports bounded conclusions

A recorded correction can show that a correction occurred.
A test result supports a claim about that defined test and scope.
Process evidence alone does not establish understanding, authorship, misconduct or a grade.
Missing trajectory evidence is missing evidence; it is not evidence of misconduct.