HE-C02 — Learning Outcomes and Assessment

Course Code: HE-C02  |  School: School of Higher Education  |  Cluster: Level 2 — Curriculum and Teaching Practice

Level: Intermediate  |  Duration: 6 weeks · 24–30 learning hours  |  Language: English  |  Certificate: Professional Certificate (non-degree)  |  Format: Self-paced with AI support under human supervision

Overview

Assessment is where a university’s claims are tested. Everything else — the syllabus, the teaching, the graduate attributes — is a promise; the assessment is the evidence. Yet assessment is routinely designed last, by individuals working alone, under time pressure, and judged by whether it can be marked quickly rather than by whether it can support the inference the degree implies.

This course rebuilds assessment from the outcome backwards. It covers the specification of outcomes that can be evidenced, the selection of task types that generate that evidence, criterion-referenced marking and rubric construction, moderation and reliability, feedback as a designed intervention, and the standard-setting practices that make a pass mark defensible rather than conventional.

Assessment security under generative AI is treated as a design question. The course argues that tasks whose validity collapses when a model is available were usually weak tasks, and it develops alternatives: process evidence, oral defence, authentic professional artefacts, and programmatic judgement across multiple points. Learners rebuild the assessment regime of a real unit and defend it against both validity and feasibility objections.

Learning outcomes

On completion, a successful learner will be able to:

1. Specify assessable learning outcomes and select task types capable of generating evidence for each.
2. Construct analytic and holistic rubrics with criteria that discriminate between levels of performance in observable terms.
3. Apply criterion-referenced standard setting and justify a pass threshold on grounds other than convention.
4. Design and run moderation procedures that produce demonstrable marking consistency across multiple markers.
5. Design feedback interventions that are timely, specific and structured so that students use them before the next assessed task.
6. Evaluate the validity, reliability, authenticity and feasibility of an assessment regime, and state the trade-offs made between them.
7. Redesign an assessment regime that remains valid in the presence of generative AI without relying primarily on surveillance.

Who this course is for

The course is for academics with responsibility for assessment design, programme and module leaders, examination and assessment officers, external examiners, quality assurance staff, and academic developers. It is discipline-neutral and applies to written, practical, performance and portfolio assessment.

Prerequisites

Completion of HE-F03 or equivalent teaching experience. Learners need access to a real unit’s assessment tasks, criteria and, where possible, anonymised marked samples.

Syllabus

Module 1 — What assessment is evidence of

The module begins with validity: the question of whether an assessment supports the inference drawn from it. It covers construct definition, construct-irrelevant variance and construct under-representation, the difference between assessing knowledge and assessing capability, and the common failure in which a task measures writing fluency or time management rather than the intended construct.

Lessons. 1.1 Validity as argument. 1.2 Construct definition and threats. 1.3 What tasks actually measure. 1.4 Auditing an existing task for validity.

Deliverable. Diagnostic note (800 words) analysing what one existing assessment actually measures.

Module 2 — Task design across modes

This module surveys task types and matches them to constructs: examinations and their legitimate uses, essays and structured written work, problem sets, practicals and laboratory reports, portfolios, presentations and oral examination, group tasks and the assessment of individual contribution within them, and authentic professional artefacts.

Lessons. 2.1 Selecting a task type for a construct. 2.2 Authentic and professional tasks. 2.3 Group work and individual attribution. 2.4 Oral assessment and defence.

Deliverable. Analytical brief (1,000 words) justifying task selection for a set of outcomes.

Module 3 — Criteria, rubrics and standards

The module covers the construction of marking instruments: analytic versus holistic rubrics, criterion wording that discriminates rather than describes, the number of levels and their descriptors, exemplar use with students, and criterion-referenced standard setting including the reasoning behind a defensible pass mark.

Lessons. 3.1 Analytic and holistic instruments. 3.2 Writing discriminating criteria. 3.3 Exemplars and shared standards. 3.4 Standard setting and the pass mark.

Deliverable. Design artefact: a complete rubric with exemplar-annotated standards.

Module 4 — Reliability, moderation and fairness

This module addresses consistency. It covers sources of marker variation, calibration and moderation procedures, double and sample marking, the treatment of borderline cases, bias in marking including the effects of name and language background, and reasonable adjustment for disability without compromising the assessed construct.

Lessons. 4.1 Sources of marker variation. 4.2 Calibration and moderation in practice. 4.3 Bias in marking and its mitigation. 4.4 Adjustment, equity and construct integrity.

Deliverable. Evaluation report (1,000 words) on marking consistency evidence in one unit.

Module 5 — Feedback as designed intervention

The module treats feedback as something students do rather than something staff produce. It covers timing relative to the next task, specificity and actionability, feedback literacy and how it is developed, self and peer assessment as feedback mechanisms, and the evidence on which feedback practices actually change subsequent performance.

Lessons. 5.1 Feedback that arrives in time to matter. 5.2 Specificity, standards and next steps. 5.3 Developing student feedback literacy. 5.4 Peer and self assessment done rigorously.

Deliverable. Implementation plan (1,000 words) for a feedback redesign with workload analysis.

Module 6 — Assessment security and generative AI

The final module confronts the current disruption. It examines which task types retain validity when a capable model is available, the evidence on detection tool accuracy and the serious harm of false accusation, process and provenance evidence, viva and defence as verification, and programmatic approaches that reduce the weight borne by any single task.

Lessons. 6.1 Which tasks survive and why. 6.2 Detection tools: accuracy and harm. 6.3 Process evidence and provenance. 6.4 Programmatic redundancy as security.

Deliverable. Final capstone: redesigned assessment regime (3,000 words) with validity and feasibility defence.

Assessment

Pass mark 70 per cent. All assessed components must be attempted. Every mark in this course is issued by a human assessor; no assessment outcome is generated automatically.

Rubric criteria

Each assessed artefact is marked against four criteria at four levels (distinction, pass with merit, pass, fail).

1. Validity reasoning. Whether the learner articulates what each task is evidence of and defends that claim against identified threats.
2. Instrument quality. Whether rubrics discriminate between levels in observable terms and whether standards are anchored to exemplars rather than adjectives.
3. Reliability and fairness. Whether moderation, bias mitigation and adjustment are addressed with concrete procedures rather than statements of intent.
4. Feasibility. Whether the regime can be delivered and marked within realistic staff workload, and whether trade-offs against validity are stated openly.

Reading list

Core. Royce Sadler, ‘Formative Assessment and the Design of Instructional Systems’, Instructional Science, 18:2 (1989), pp. 119–144.

David Boud & Nancy Falchikov (eds), Rethinking Assessment in Higher Education: Learning for the Longer Term (London: Routledge, 2007).

Cees P. M. van der Vleuten et al., ‘A Model for Programmatic Assessment Fit for Purpose’, Medical Teacher, 34:3 (2012), pp. 205–214.

David Carless & David Boud, ‘The Development of Student Feedback Literacy’, Assessment & Evaluation in Higher Education, 43:8 (2018), pp. 1315–1325.

Michael Kane, ‘Validating the Interpretations and Uses of Test Scores’, Journal of Educational Measurement, 50:1 (2013), pp. 1–73.

Sally Brown & Peter Knight, Assessing Learners in Higher Education (London: Kogan Page, 1994).

Academic integrity and use of AI

AI tools may be used in this course for locating literature, organising evidence, drafting non-analytical sections and criticising the learner’s own work. They may not be used to generate the analysis, the evaluation or the recommendation. Every submission carries a use-of-AI statement naming the tools used and the purpose of each use.

All citations must be verified against the primary source, and institutional facts — regulations, standards, accreditation criteria, published data — must be cited to the issuing body. Fabricated references and invented statistics result in failure and referral. Where learners write about their own institution they must not disclose confidential personnel or student data, and illustrative cases must be anonymised.