EAQE cheat sheet

A quality framework is a structured set of principles, criteria and processes that defines what quality means and how it is evidenced, judged and improved within an institution or system.

Standards are normative statements of expected performance; criteria are the operational dimensions used to judge whether a standard is met; indicators are the measurable signals of performance against a criterion.

Common framework architectures include: (a) fitness-for-purpose, judging quality against stated institutional mission; (b) standards-based, judging against external threshold criteria; (c) value-added, judging the learning gain from entry to exit.

Typical framework components: scope and definitions, standards and criteria, evidence requirements, self-evaluation process, external review method, decision categories and reporting conventions.

The ESG (Standards and Guidelines for Quality Assurance in the European Higher Education Area) structure has three parts: internal quality assurance, external quality assurance, and quality assurance agencies themselves.

The PDCA (Plan-Do-Check-Act) cycle, derived from Deming, is the most common improvement logic embedded in frameworks: plan objectives, implement, monitor against indicators, and act on findings.

A framework's decision categories are usually graded (e.g., full confidence, limited confidence, no confidence) or binary (meets/does not meet); the category must map to defined consequences such as accreditation duration.

Alignment means each standard is traceable to a criterion, an indicator, an evidence source and a review question; a standard with no evidence route is unassessable and must be revised.

Governance is the allocation of decision rights and accountability; quality governance specifies who sets standards, who monitors performance, and who acts on findings.

Separation-of-duties rule: the body that sets quality policy should not be the sole body that judges compliance with it; internal audit or external review provides independent assurance.

A quality council or senate committee typically owns approval of quality policy; faculties or programme teams own implementation; an independent quality unit owns methodology and data integrity.

Strategic quality management links institutional mission and strategy to measurable objectives via a quality policy, a multi-year improvement plan, and annual operational targets.

Board-level oversight questions: is there a documented quality policy, are roles and responsibilities defined, is performance reported to the governing body, and are findings closed out with evidence?

Risk-based governance prioritises assurance effort on programmes or processes with high impact and low control (e.g., new programmes, high-failure modules, placements and partnerships).

Leadership for quality is evidenced by resource allocation, not statements: staffing of the quality function, data systems, staff development budgets and time protected for review cycles.

Policy lifecycle rule: every quality policy needs an owner, an approval date, a review interval (commonly 3-5 years), and a version-controlled change log.

Internal quality assurance (IQA) is the institution's own system of planning, monitoring, reviewing and improving quality; it is the primary locus of assurance, with external review providing validation.

A self-evaluation (self-assessment) report is a structured, evidence-referenced narrative against each standard, stating strengths, weaknesses, and improvement actions with owners and deadlines.

Effective self-evaluation uses triangulated evidence: documentary evidence, quantitative indicator data, and stakeholder testimony (students, staff, employers, graduates).

Programme review cycles are typically 4-7 years; annual monitoring covers enrolment, progression, completion, grade distributions, student feedback and action closure.

A self-evaluation must distinguish description from judgement: statements of activity are not findings; each claim of strength requires evidence and each weakness requires an improvement action.

Internal audit or a 'critical friend' review before external evaluation reduces the risk of over-claiming and identifies missing evidence trails.

Evidence register rule: each standard maps to a named evidence item, its location, its owner and its date; evidence older than the review period must be justified.

Action-tracking discipline: improvement plans record action, owner, deadline, resource, success measure and closure evidence; open actions carried over more than two cycles indicate systemic weakness.

External quality assurance (EQA) is independent evaluation of an institution or programme against published standards, conducted by an agency outside the unit being reviewed.

Main EQA methods: accreditation (judgement of threshold compliance), audit (evaluation of the institution's own quality system), assessment/evaluation (judgement of quality level), and recognition/listing.

A typical external review sequence: application and self-evaluation submission, evidence review, site visit with stakeholder meetings and observation, draft report with factual check, decision, publication, follow-up.

Peer review panels combine subject expertise, quality-assurance expertise, and often a student member; panel members must declare conflicts of interest and sign confidentiality undertakings.

Judgements must be traceable: each finding cites the standard, the evidence reviewed, and the reasoning; findings not grounded in evidence are open to appeal.

Appeal and complaint channels are procedural safeguards: appeals address procedural error or factual inaccuracy, not disagreement with professional judgement.

Accreditation is time-limited; renewal depends on evidence of sustained compliance and closure of prior conditions, not merely on the absence of new problems.

Mutual recognition and cross-border arrangements rest on comparability of standards and reviewer competence, not on automatic equivalence of outcomes.

Stakeholder mapping rule: identify each group's interest and influence, then choose the mechanism - information, consultation, participation in decision-making, or partnership.

Core stakeholder groups: students, teaching and administrative staff, employers and professional bodies, graduates/alumni, funders and regulators, and wider community.

Student-centred quality means the design of provision starts from intended learning outcomes and student needs; students are partners in design, review and improvement, not only respondents to surveys.

Student representation mechanisms include course representatives, student members on quality committees and review panels, and structured student submission to external review.

Survey metrics commonly used: satisfaction, engagement, workload, assessment and feedback, and belonging; response rate and representativeness must be reported alongside scores.

Closing the feedback loop is the test of engagement: students must be told what changed as a result of their input, with a named owner and a date.

Employer and professional-body input is most valid at the level of programme outcomes and graduate competence, not at the level of individual module delivery.

Participation quality rule: consultation without influence on decisions is tokenism; document the decision, the evidence considered, and the reason for the outcome.

An indicator is a quantitative or qualitative signal of performance; a KPI is an indicator tied to a target and an accountable owner.

Indicator quality criteria: valid (measures what it claims), reliable (consistent across time and units), comparable (defined identically across units), timely, and actionable.

Common education indicators: student-staff ratio, progression and completion rates, attrition, grade distribution, graduate employment, satisfaction scores, staff qualifications, research output.

Rate formulas: progression rate = students progressing to next stage / students eligible to progress; completion rate = qualifiers within the reference period / entrants to the cohort; attrition = 1 - completion for the same cohort.

Cohort definition is critical: a completion rate is only comparable if the entrant cohort, the tracking window and the treatment of transfers and withdrawals are identical.

Small-number caution: with fewer than about 30 cases, percentages are unstable; report counts alongside percentages and suppress or flag cells that risk identifying individuals.

Data governance: define each indicator once in a data dictionary (name, formula, source, owner, refresh cycle), and manage version changes to preserve trend comparability.

Triangulation rule: no consequential decision should rest on a single indicator; combine quantitative trends with documentary and qualitative evidence.

Correlation is not causation; a rise in satisfaction alongside a grade increase may reflect assessment design, cohort composition or marking practice rather than teaching quality.

Continuous improvement is the systematic, cyclical use of evidence to close performance gaps; it requires a baseline, a target, an intervention and a re-measurement.

PDCA and its variants (Define-Measure-Analyse-Improve-Control in Six Sigma; Plan-Do-Study-Act in improvement science) share one logic: small, tested changes with measurement before scaling.

Root-cause tools: the '5 Whys' and cause-and-effect (fishbone) analysis distinguish symptoms from causes; interventions aimed at symptoms produce recurrence.

Change models: Kotter's eight steps (urgency, coalition, vision, communication, enabling action, short-term wins, consolidation, anchoring) and Lewin's unfreeze-change-refreeze.

Readiness for change depends on perceived benefit, clarity of the change, capability to implement, and available resources; neglect of any one produces resistance or drift.

Improvement plan elements: gap statement, root cause, action, owner, resources, timeline, success measure, review point; plans without success measures cannot be evaluated.

Anchoring rule: a change is sustained only when it is embedded in routine processes, job descriptions, induction, data systems and policy - otherwise performance regresses.

Maturity models (e.g., reactive, planned, managed, optimised) allow an institution to judge whether improvement is ad hoc or systemic.

Constructive alignment: intended learning outcomes, teaching and learning activities, and assessment tasks must all target the same cognitive level and content.

Learning outcomes should be observable and assessable, expressed with action verbs mapped to a taxonomy level (e.g., remember, apply, evaluate, create) rather than vague verbs such as 'understand'.

Assessment validity is the degree to which an assessment measures the intended outcomes; reliability is the consistency of results across markers, occasions and items.

Assessment types: formative (feedback for learning, low or no stakes) and summative (judgement of achievement); diagnostic assessment establishes starting points.

Marking quality tools: rubrics or criteria with descriptors, anonymised double marking, moderation and external examiner oversight; inter-rater agreement indicates reliability.

Feedback quality criteria: timely, specific, criterion-referenced, actionable, and forward-looking; feedback that only justifies a grade does not improve learning.

Workload and credit rules: one credit typically represents 25-30 hours of total student effort; module workload must be consistent with its credit value.

Academic integrity controls: assessment design that reduces opportunity (authentic tasks, in-person components), clear policy, detection tools, and proportionate sanctions.

Teaching quality evidence includes peer observation, student feedback, qualification and development records, and learning analytics on engagement, not only satisfaction scores.

Core ethical principles in quality practice: integrity of evidence, impartiality of judgement, confidentiality of personal data, fairness to those judged, and avoidance of conflicts of interest.

Conflict-of-interest rule: reviewers must declare any relationship (employment, supervision, recent collaboration, commercial interest) with the unit reviewed and recuse where a reasonable observer would doubt impartiality.

Transparency means published standards, published review methods, published outcomes (with appropriate redaction), and reasons for decisions that affected parties can scrutinise.

Accountability requires a named owner for each quality decision and each improvement action, plus a record of the evidence on which the decision was based.

Data ethics: collect the minimum necessary, obtain informed consent where required, anonymise before analysis where identity is not needed, and apply retention limits.

Evidence integrity: fabricating, selectively presenting, or suppressing inconvenient evidence is misconduct; self-evaluation must report weaknesses as well as strengths.

Fairness in judgement: the unit being evaluated must have the opportunity to correct factual errors before publication (factual accuracy check).

Student and staff data used in quality processes are subject to data-protection law; publication of small-cell results can enable re-identification and must be suppressed.

Quality culture comprises shared values, beliefs and behaviours about quality, plus the structures and processes that support them; it is assessed through both.

Indicator behaviours of a strong quality culture: staff can explain quality processes, findings are discussed openly, improvement is resourced, and students participate in decisions.

Compliance culture versus improvement culture: compliance meets external requirements with minimum effort; improvement uses evidence to change practice, with compliance as a by-product.

Institutional learning occurs when findings from reviews, audits and incidents are analysed collectively for patterns and fed into policy and strategy, not filed per unit.

Psychological safety - the belief that raising problems will not be punished - is a precondition for honest self-evaluation; blame-oriented responses suppress evidence.

Staff development in quality literacy (writing outcomes, using data, self-evaluation) is a leading indicator of quality maturity.

Communication rule: quality findings must reach the people who can act on them, in a form they can use - short, specific, and linked to their own practice.

Culture change is measured by behaviour and decisions over time, not by declarations; repeated cycles showing closed actions and changed practice are the evidence.