# Language Programme Review Framework

A programme review asks: is the programme achieving what it promises, for whom, and what should change? It
draws on evidence from several sources and ends in prioritised, owned recommendations.

## Review areas

Based on the areas used by language-school quality schemes (Eaquals, British Council Accreditation UK, CEA,
NEAS, ALTO and similar):

| Area | Questions | Evidence |
|------|-----------|----------|
| Aims and design | Are outcomes clear, level-referenced and matched to learner needs? | Syllabus, can-do statements, needs analysis |
| Curriculum and materials | Is the syllabus coherent, sequenced, recycled? Are materials appropriate and current? | Scope and sequence, materials audit |
| Teaching and learning | Is teaching effective and learner-centred? | Observation records, learner work, learner feedback |
| Assessment | Placement accurate? Progress tracked? Exit tests valid, reliable, aligned? | Placement data, test analysis, moderation records |
| Learner outcomes | Do learners progress, pass, stay, achieve goals? | Retention, level gain, pass rates, destinations |
| Learner support and welfare | Orientation, tutorials, counselling, safeguarding, accessibility | Policies, records, interviews |
| Staffing and development | Qualifications, observation, CPD, workload | Staff records, CPD log, observation cycle |
| Resources and environment | Rooms, technology, self-access, online platform | Audit, learner and staff feedback |
| Management and quality | Planning, data use, complaints, continuous improvement | Policies, minutes, action plans |
| Marketing and information | Is published information accurate (levels, hours, outcomes)? | Website, brochures vs reality |

## Evidence sources

| Source | Method | Watch for |
|--------|--------|-----------|
| Learner data | `program_metrics.py` on enrolment, status, levels, results, attendance | Small groups; missing data; who withdrew and why |
| Learner voice | Surveys, focus groups, exit interviews | Survey fatigue; only satisfied learners respond |
| Teacher voice | Interviews, staff survey | Power dynamics; anonymity |
| Observation | Sample of lessons against agreed criteria | One-off lessons; observer bias |
| Documents | Syllabus, plans, tests, policies | Documents that do not match practice |
| Stakeholders | Employers, sponsors, partner institutions | Their goals may differ from learners' |
| Benchmarks | Other programmes, previous years, targets | Different intakes and definitions |

Triangulate: a finding needs at least two sources.

## Key indicators

| Indicator | Definition | Typical concern level |
|-----------|------------|-----------------------|
| Retention | Completed / (completed + withdrawn) | Under 75-80% |
| Pass rate | Completers meeting exit criteria | Under 75% or far above (test too easy?) |
| Level gain | Median CEFR half-steps per course | Zero or below the hours-based expectation |
| Attendance | Mean % of lessons attended | Under 80% |
| Satisfaction | Mean rating 1-5 | Under 3.5, or a drop from last year |
| Placement accuracy | Learners moved level in first two weeks | Over 10% |
| Teacher observation | % lessons meeting criteria | Trends, not single scores |

Compare against expected gains from guided hours (see the cefr-aligned-syllabus-builder skill): a 60-hour course
rarely moves a learner a full CEFR level at B1+.

## From findings to recommendations

| Finding (evidence) | Cause (judgement) | Recommendation | Priority | Owner | By |
|--------------------|-------------------|----------------|----------|-------|----|
| Evening B1 retention 62%; exit interviews cite timing | Class ends 21:30; working adults | Pilot 2x2h instead of 3x1.5h; online option | High | Programme director | Term 2 |

Priorities: **high** = learner outcomes or safety at risk; **medium** = quality or efficiency; **low** = good to
have. Every recommendation is specific, owned, dated and measurable.

## Ethics and data

- Anonymise learner data; report groups under 10 with caution or not at all.
- Share the draft with staff before publication; correct factual errors, keep judgements.
- Separate review of the programme from appraisal of individual teachers.