Turning Course Evaluation Data into ABET Accreditation Evidence (Criterion 3 & 4)

Short answer: ABET does not require a specific course-evaluation tool, but it does require that you regularly use documented processes to assess and evaluate student-outcome attainment, and systematically use the results for continuous improvement (Criterion 4). Course evaluations and student surveys are indirect measures of outcome attainment — they capture students'' perceptions of what they learned, which complement direct measures (exams, rubric-scored projects). The challenge most programmes face is not collecting the data; it is turning a pile of Likert scores and free-text comments into evidence of evaluation and action that a review team will accept. This guide maps ABET''s expectations to concrete, audit-ready outputs, and shows how an AI-moderated interview platform like Koji for Education produces the standardised, longitudinal, action-linked evidence that the continuous-improvement criterion rewards.

Who this is for: programme chairs, ABET coordinators, assessment committees, and quality/IR leaders at engineering, computing, engineering-technology, and applied-science programmes — including the growing number of internationally oriented European institutions seeking ABET recognition alongside or instead of EUR-ACE.

ABET in brief, and why feedback data matters

ABET (the Accreditation Board for Engineering and Technology) accredits programmes in applied and natural science, computing, engineering, and engineering technology. As of publication, ABET reports that 4,863 programmes at 950 colleges and universities across 42 countries have earned accreditation, with over 200,000 students graduating from accredited programmes each year. Accreditation is programme-level and outcomes-based, which makes evidence of how you measure and improve central to a successful review.

Two criteria do most of the work for evaluation data:

• Criterion 3 — Student Outcomes: the programme must have documented student outcomes (for engineering, the well-known seven outcomes 1–7, covering problem-solving, design, communication, ethical and professional responsibility, teamwork, experimentation and data analysis, and the ability to acquire and apply new knowledge).
• Criterion 4 — Continuous Improvement: the programme must regularly use appropriate, documented processes for assessing and evaluating the extent to which the student outcomes are being attained, and the results must be systematically utilised as input for the programme''s continuous improvement actions.

ABET draws a precise distinction that matters for how you use survey data:

• Assessment is the process of identifying, collecting, and preparing data. Effective assessment uses relevant direct, indirect, quantitative, and qualitative measures as appropriate to the outcome.
• Evaluation is the process of interpreting that data to judge the extent of outcome attainment, resulting in decisions and actions.

Course evaluations sit squarely in the indirect-assessment category, alongside exit surveys, focus groups, and interviews. They are not a substitute for direct measures, but they are explicitly recognised, and they are uniquely good at one thing direct measures cannot do: explaining why attainment is where it is, and what students would change.

The real problem: from comments to evidence

A standard end-of-course survey produces two artefacts: numeric ratings and a column of free-text comments. For ABET, neither is evidence on its own. A reviewer wants to see a closed loop: data collected → interpreted (evaluation) → a decision made → an action taken → re-measured next cycle. The free-text comments are usually where the actionable insight lives, but coding hundreds of comments per cohort into outcome-aligned themes is exactly the manual labour most programmes lack the staff to sustain. So the qualitative evidence either goes unused or gets summarised so loosely that it carries no weight in the self-study.

This is where the method of collection changes what evidence you can produce.

Requirement → Koji output mapping

The table below maps common ABET evidence expectations to concrete outputs an AI-moderated interview platform can generate. The same AI interview engine that powers customer and user research on the main Koji platform is applied to course and programme evaluation in Koji for Education.

ABET expectation	What reviewers want to see	Koji output that supports it
Criterion 4 — documented assessment process	A repeatable, standardised collection method	Standardised AI-moderated interview protocol applied identically across sections and cohorts
Indirect measure of student outcomes (Crit. 3 & 4)	Outcome-aligned student perception data	Interviews mapped to specific student outcomes; themes tagged per outcome
Evaluation (interpretation), not just data	Themed interpretation, not raw exports	Automatic thematic analysis clustering responses into themes with verbatim quotes as evidence
Continuous improvement loop	Decision → action → re-measurement	Themes-to-actions tracking carried across terms with status
Longitudinal trend evidence	Multi-cohort comparison over time	Cohort-over-cohort theme and sentiment reporting
Consistency / reduced rater bias	Standardised, neutral probing	Bias-aware, standardised moderation — every student probed the same neutral way
Audit-ready documentation	Exportable, attributable records	Exportable reports linking findings to verbatim evidence and to the action taken

Koji supports the indirect-assessment and continuous-improvement portions of an ABET self-study. It does not replace direct measures of outcome attainment (e.g., rubric-scored coursework or exam analysis), which remain essential.

A practical continuous-improvement cycle

1. Define the question per outcome. Instead of one generic "rate this course," configure short interviews that probe specific outcomes — e.g., for Outcome 5 (teamwork), ask students to describe a concrete team experience and what helped or hindered it.
2. Collect with standardised probing. The AI moderator asks neutral follow-ups when answers are vague ("you said the project was disorganised — what specifically was unclear?"), producing attributable, specific data rather than one-line complaints.
3. Evaluate automatically. Thematic analysis clusters responses into outcome-aligned themes, each backed by verbatim quotes — the interpretation step ABET distinguishes from raw assessment.
4. Decide and act. The committee records a decision against a theme (e.g., restructure the team-project brief) as a tracked action.
5. Re-measure next cycle. The following cohort''s interviews show whether the action moved the theme — the longitudinal evidence that closes the loop.

This sequence is precisely the "documented process → systematic use of results → continuous improvement" narrative Criterion 4 asks programmes to demonstrate.

When a dedicated assessment-management tool is the better choice

Honesty matters with a review-literate audience, so be clear about scope. If your primary need is to manage the full ABET assessment lifecycle — curriculum mapping, direct-measure rubric scoring tied to outcomes, assessment plans, and self-study document assembly — then a dedicated assessment-management platform (for example, Watermark''s assessment suite, Anthology, or similar) is purpose-built for that, and an interview tool does not replace it. Likewise, your direct measures of outcome attainment must come from coursework and examinations, not from perception surveys.

Koji''s contribution is specific and complementary: it is the strongest option when your weak point is the indirect, qualitative half of the evidence base — turning what students actually experienced into standardised, themed, action-linked documentation without a coding team. Many programmes pair the two: a direct-measure/assessment-management system for outcome scoring and self-study assembly, and Koji for the indirect, formative evidence that explains the why and demonstrates closing the loop.

European context: ABET alongside EUR-ACE

For European engineering schools, ABET is increasingly pursued in addition to European routes such as the EUR-ACE label. The good news is that the evidence model is portable: standardised, longitudinal, action-linked student feedback satisfies both ABET''s Criterion 4 continuous-improvement narrative and the ESG-aligned "quality cycle" expectations of European agencies. Collecting it once, in a GDPR-first EU-data-handling platform, lets a programme feed two accreditation stories from a single evidence base.

Related Resources

• Course Evaluation Evidence for EUR-ACE Engineering Accreditation
• Course Evaluation Evidence for AACSB & EQUIS
• Course Evaluation Evidence for WFME Medical-School Accreditation
• Course Evaluation as NVAO Accreditation Evidence
• Student Feedback as ESG / ENQA Accreditation Evidence

Ready to build an ABET-ready evidence base from student feedback? Explore Koji for Education or book a walkthrough.

A sample ABET evidence package

When a review team examines your continuous-improvement process for a given outcome, a strong evidence package built from interview data looks like this:

1. The protocol. A one-page description of the standardised interview used each term, the outcomes it targets, and how it is administered — demonstrating a documented, repeatable process (Criterion 4).
2. The assessment data. The collected interviews for the cohort, anonymised, retained in a GDPR-first EU environment.
3. The evaluation artefact. The thematic-analysis report: outcome-aligned themes, each with frequency and verbatim quotes — showing that data was interpreted, not merely exported.
4. The decision record. Minutes or a tracked log showing the assessment committee reviewed the themes and resolved on a specific action (e.g., "revise lab brief for Outcome 6").
5. The re-measurement. The following cohort''s report against the same theme, showing whether the action moved the needle.

A reviewer reading that sequence sees a closed loop, not a data dump — which is the single most common gap flagged in continuous-improvement findings.

Common pitfalls to avoid

• Treating perception data as a direct measure. Indirect measures support, but never substitute for, rubric-scored coursework and exam analysis. Keep both.
• Collecting but not acting. Data with no documented decision-and-action trail is the classic Criterion 4 weakness. Track actions, not just results.
• Resetting every cycle. The value is longitudinal; preserve theme continuity across cohorts so trends are visible.
• Over-claiming. Frame student feedback precisely as indirect evidence of perceived attainment — reviewers trust honest, well-scoped claims far more than inflated ones.