Assurance of Learning (AoL) is the part of accreditation that proves, with evidence, that a program delivers the learning it promises. It is a distinct workload from designing a single course: it operates at the degree-program level, ties program learning outcomes (PLOs) to the curriculum through a mapping matrix, gathers direct and indirect assessment evidence against those outcomes, and writes the closing-the-loop narrative showing the program changed something in response to the data.
This skill scaffolds that documentation for the two accreditors AlterLab faculty most often face — AACSB (business) and ABET (engineering & computing) — in each accreditor's own structure. It drafts text and checks the shape of an outcome-mapping matrix; the school still files its self-study through the accreditor's own portal.
When to Use This Skill
Use this skill when the user needs to:
- Write or refine a set of program learning outcomes / competency goals for a degree program (AACSB competencies, ABET student outcomes).
- Build a curriculum-to-outcome mapping matrix (which courses Introduce / Reinforce / Master each outcome) and check it for coverage gaps.
- Design a direct + indirect assessment plan against the outcomes (which measure, in which course, on what cycle).
- Draft rubrics keyed to a program outcome for direct assessment of student work (the assessment instrument, not a course grade).
- Write a closing-the-loop / continuous-improvement narrative: results → interpretation → action → re-assessment.
- Draft AACSB Standard 5 or ABET Criterion 3/4 self-study prose, or an AoL section of a continuous-improvement report.
Does NOT Trigger — route these elsewhere
| The request is really about… | Route to |
|---|---|
| Designing a single course — its schedule, activities, course-level rubric, learning objectives for one class | alterlab-teaching-design |
| Writing an AI-use / academic-integrity policy for a syllabus | alterlab-syllabus-ai-policy |
| Post-award grant progress/final reports, RPPR, periodic reports, budget variance | alterlab-grant-reporting |
| Designing a survey or questionnaire instrument as a research tool (validity, scales) | alterlab-survey-design |
| An individual's doçentlik promotion eligibility against ÜAK criteria | alterlab-docentlik-eligibility |
| YÖKATLAS program placement / quota / ranking data for a Turkish program | alterlab-yokatlas |
| Verifying the citations in a self-study actually exist | alterlab-citation-verifier |
If the work is about one course rather than the whole program, stop and route to
alterlab-teaching-design. This skill is program-level only.
The two accreditors (verified)
Do not invent standard numbers, outcomes, or criterion text. The verbatim
outcome lists and the full per-standard breakdown live in
references/accreditor_standards.md. The
essentials:
AACSB — 2020 Business Accreditation Standards, Standard 5: Assurance of Learning
AoL sits in the curriculum cluster of the 2020 standards: Standard 4 Curriculum, Standard 5 Assurance of Learning, Standard 6 Learner Progression, Standard 7 Teaching Effectiveness and Impact. Standard 5 has four sub-standards:
| § | Title |
|---|---|
| 5.1 | Assurance of Learning Processes (direct + indirect measures; results lead to curricular and process improvements) |
| 5.2 | Degree Equivalency (equivalent outcomes across location/modality) |
| 5.3 | Stackable Microlearning Credentials |
| 5.4 | Non-Degree Executive Education |
Standard 5 is principles-based. AACSB's own position is that more complexity, more competency goals, or assessing every competency every year does not make an AoL process better — a systematic, mission-informed process that produces meaningful improvement does. Key requirements (5.1): competencies are identified per degree program, derive from the school's mission, are reported at the degree level (not the major level), and AoL must employ both direct and indirect measures across the assessment portfolio.
- Direct measures — evidence from learner work based on direct observation of performance: examinations, quizzes, assignments, internship/externship feedback.
- Indirect measures — third-party input not based on direct observation: e.g. employer surveys, focus groups, interviews, external outcome measures.
AACSB AoL is concerned with broad, program-level competency goals for each
degree program, not detailed goals by course or topic — that course-level work
is alterlab-teaching-design's job.
Which AACSB edition applies? AACSB adopted the 2026 Global Standards for
Business Education (ratified April 2026, effective 1 July 2026). 2026–27 is a
pilot year: about 30 volunteer schools are reviewed under the new standards while
the rest continue under the 2020 standards (a CIR visit in 2026–27 may use
either); the 2026 standards apply to everyone from 2027–28. Standard 5 keeps its
title and its 5.1–5.4 structure, but 5.1 now expects learning objectives
derived from each competency, ties direct and indirect measures to those
competencies/objectives "as opposed to simple satisfaction measures", lists
curriculum maps in the suggested documentation, and requires Table 5-1 only
for initial accreditation. Ask which edition the school's next review uses before
drafting — the side-by-side is in
references/accreditor_standards.md.
ABET — Criterion 3 (Student Outcomes) + Criterion 4 (Continuous Improvement)
ABET frames AoL across linked criteria: Criterion 2 Program Educational Objectives (what graduates achieve a few years out), Criterion 3 Student Outcomes (what students can do by graduation), Criterion 4 Continuous Improvement (the assessment/evaluation loop that uses outcome data to improve the program).
ABET's engineering criteria specify seven required student outcomes (1)–(7) — problem-solving, engineering design, communication, ethical/professional responsibility, teamwork, experimentation & data analysis, and acquiring/applying new knowledge. The verbatim text of all seven is in the references file (checked against the 2026–2027 EAC criteria, where they are unchanged); quote them exactly rather than paraphrasing a numbered outcome into the self-study, because reviewers check the program's outcomes against the published wording. ABET republishes its criteria every cycle, so confirm the edition for the review year.
Computing programs (CAC) and other commissions use a differently-worded outcome set. This skill scaffolds the engineering (EAC) seven outcomes by default; confirm the commission and pull that commission's exact outcome text before drafting. If you cannot confirm the exact wording, say so rather than invent it.
How to build an AoL package
- Establish scope first. Which accreditor (AACSB vs ABET), which commission, and which degree program? AoL is per-program. If unstated, ask before drafting — the structure differs by accreditor.
- Fix the outcomes. For ABET, use that commission's exact published student outcomes (engineering = the seven in the references file). For AACSB, define program competency goals derived from the school's mission, stated at the degree level. Each outcome must be observably assessable.
- Map the curriculum to the outcomes. Build a matrix: rows = required courses, columns = outcomes, cells = coverage level (commonly Introduced / Reinforced / Mastered, or Assessed). Every outcome needs at least one course where it is assessed; flag any outcome with no assessment point. Validate the matrix structure with the script below.
- Plan direct + indirect assessment. For each outcome pick at least one direct measure (embedded exam question, rubric-scored assignment, project, internship evaluation) and round out the portfolio with indirect measures (employer/alumni survey, exit interview, focus group). Record measure, course/point of collection, target/benchmark, and cycle. AACSB requires both measure types across the portfolio; do not present indirect evidence alone for the whole school.
- Write rubrics keyed to outcomes (not to a course grade). Each row is a
trait tied to the outcome; levels describe observable performance. Templates:
references/aol_templates.md. - Close the loop. Report results against the target, interpret them,
state the action taken (curriculum/process change), and the re-assessment
point. A loop that reports data but changes nothing is not closed — both
accreditors look for the change. Detail and worked narrative:
references/closing_the_loop.md. - Flag, never fabricate. Assessment results, cohort sizes, target
percentages, course numbers, and dates come only from the user's materials.
Mark anything missing as
[TODO: confirm]. Never invent an outcome score, a standard number, or an outcome's wording. Hand any reference list toalterlab-citation-verifier.
Validate the mapping matrix — don't eyeball it
A coverage gap (an outcome no course assesses) or a thin matrix (an outcome touched only at "Introduced") is exactly what a peer-review team flags. Check the matrix structurally instead of reading it by hand:
uv run python skills/faculty-life/alterlab-accreditation-aol/scripts/aol_matrix.py \
path/to/matrix.csv --assessed-levels M,A --json
The CSV has a course column and one column per outcome; cells hold coverage
codes (e.g. I, R, M, A, or blank). The script reports, per outcome:
whether it is covered at all, whether it has at least one assessment-level
cell (default M or A), and the count by level — then a program-level verdict
(OK / GAPS). It does no network I/O and uses only the Python standard
library. It checks structure and coverage, not whether the pedagogy is sound —
that judgment stays with the user and the faculty.
See references/aol_templates.md for the matrix
CSV layout, the I/R/M scheme, and example rows.
Guardrails
- Program-level, not course-level. This skill never writes a course
schedule, course-grade rubric, or single-class objectives — that is
alterlab-teaching-design. It maps a program to its outcomes. - No invented standards or outcomes. Standard/criterion numbers and the exact text of student outcomes come from the references file (verified against the accreditors' published criteria). If the user's program uses a different commission, get that commission's exact outcome text first; do not paraphrase.
- No invented results. Scores, targets, cohort counts, course numbers, and
dates come from the user. Mark gaps
[TODO: confirm]. - The school submits, not the skill. This skill drafts text and checks matrix structure; the school files its self-study through AACSB myAccreditation or the ABET portal. The skill has no portal access and claims none.
- Both measure types. Don't let an AACSB portfolio rest on indirect evidence alone; pair direct and indirect measures across programs.
References
references/accreditor_standards.md— AACSB 2020 Standard 5 (5.1–5.4) detail with direct/indirect definitions, and the verbatim ABET Criterion 2/3/4 text including all seven engineering student outcomes, each with its source.references/aol_templates.md— PLO-writing guidance, the curriculum-to-outcome mapping matrix CSV layout (I/R/M scheme), a direct/indirect assessment-plan table, and an outcome-keyed rubric template.references/closing_the_loop.md— the results → interpretation → action → re-assessment structure with a worked closing-the-loop narrative and the questions each accreditor asks.scripts/aol_matrix.py— stdlib-only curriculum-to-outcome coverage validator.
Part of the AlterLab Academic Skills suite.