Appreciative Inquiry for Course Evaluation: A Strengths-Based Alternative to Deficit Surveys

BLUF: Appreciative Inquiry (AI) is a strengths-based, generative method for organizational change developed by David Cooperrider and Suresh Srivastva in 1987 — and it is not artificial intelligence. Applied to course evaluation, it deliberately inverts the deficit logic of "what went wrong?" surveys by asking students and staff to identify when a course worked at its best and how to design more of that. Used alongside (not instead of) diagnostic questions, this approach can improve response engagement, staff buy-in, and the odds that feedback actually changes teaching.

A note on the acronym

Throughout this article, "AI" means Appreciative Inquiry — the organizational-development methodology — unless we explicitly say "artificial intelligence." This matters because course-evaluation technology increasingly involves both: the appreciative-inquiry philosophy of asking generative questions, and artificial-intelligence tooling that moderates conversations and analyses responses. We keep them separate on purpose.

What the research says

Origins: Cooperrider and Srivastva, 1987

Appreciative Inquiry was introduced by David L. Cooperrider and Suresh Srivastva in their 1987 chapter "Appreciative inquiry in organizational life," published in Research in Organizational Change and Development, Volume 1, pages 129–169. Their argument was a direct challenge to the dominant "action research" paradigm of the time, which framed organizations as problems to be diagnosed and fixed. Cooperrider and Srivastva contended that this deficit orientation was itself a limiting force: if you build inquiry around what is broken, you get an organization that becomes expert at analysing its own failures while starving its capacity to imagine what it could become.

Their alternative rests on a social-constructionist premise: organizations are not fixed machines but living constructions that are continually made and re-made through language, conversation, and shared meaning. From this it follows that the questions we ask are not neutral — they are fateful. Inquiry is intervention. The very act of asking "when has this system been most alive, effective, and valued?" begins to shift the reality being studied, because it directs collective attention toward strengths, exceptions, and possibilities rather than deficits. This is often summarised as the heliotropic principle: social systems, like plants toward light, tend to grow in the direction of what they persistently study.

The 4-D cycle

Practitioners operationalise Appreciative Inquiry through a four-phase cycle, commonly rendered as Discovery, Dream, Design, and Destiny (the "4-D cycle"):

• Discovery — appreciate "the best of what is." Participants surface stories of peak experiences: the moments when the course, the teaching, or the learning community worked at its very best. The unit of analysis is the positive exception, not the average complaint.
• Dream — imagine "what might be." Building on those stories, participants articulate an aspirational image of the course at its best, freed from the constraint of present problems.
• Design — determine "what should be." The group co-constructs concrete propositions and structures — assessment redesign, sequencing, contact formats — that would make the dream reproducible rather than accidental.
• Destiny (sometimes "Deliver") — sustain "what will be." The system commits to and enacts the changes, with attention to the self-organizing momentum that keeps them alive after the intervention ends.

Generativity: the mechanism that makes it work (Bushe)

A crucial refinement comes from Gervase Bushe, who has spent two decades arguing that the "positive" framing of Appreciative Inquiry is often misunderstood as mere optimism. In Bushe and Kassam's meta-case analysis — "When is appreciative inquiry transformational? A meta-case analysis," Journal of Applied Behavioral Science, 41(2), 161–181 (DOI: 10.1177/0021886304270337) — the authors examined twenty published AI cases. Only seven (35%) produced genuinely transformational change. Tellingly, transformation did not correlate with how positive or upbeat the process was, nor with strict adherence to all five classical AI principles. What distinguished the transformational cases was generativity: the creation of new ideas, images, and metaphors that changed how people thought, making new decisions and actions available that had not occurred to them before, plus a focus on supporting self-organizing change rather than dictating specific behaviours.

Bushe's later theoretical work develops this into a caution and a design imperative: Appreciative Inquiry works not because it feels good, but when it generates new ways of seeing. A strengths-based course evaluation that merely collects compliments is not doing Appreciative Inquiry in any meaningful sense; the value lies in producing generative insight that reframes what a course could be.

Corroboration in higher education

Appreciative Inquiry has a substantial applied literature in higher education. Sacco-Bene's study — "Appreciative inquiry: A component of course evaluation and improvement," The Journal of Humanistic Counseling, 61(3), 170–183 (DOI: 10.1002/johc.12180) — reports on a cohort of counseling students who used the Discovery and Dream stages to collaboratively evaluate and improve their courses, producing aspirational statements and an action plan for future teaching staff. The account illustrates the closing-the-loop advantage: because students help construct the design, they are invested in whether it is enacted.

A broader synthesis is provided by Cho and Ardichvili's integrative review — "Appreciative Inquiry: An Integrative Review of Studies in Three Disciplines," Human Resource Development Review, 23(3), 376–401 (DOI: 10.1177/15344843241256156) — which examined the AI literature across healthcare, higher education, and management. The reviewers found generally positive effects at individual, group, and organizational levels, and, echoing Bushe and Kassam, noted that benefits could accrue even when practitioners did not follow every step of the 4-D cycle. That flexibility is useful for course evaluation, where a full multi-day summit is rarely feasible.

Why it matters for course evaluation in practice

Most Student Evaluation of Teaching (SET) instruments are, structurally, deficit-hunting devices. Their Likert items and open-text prompts overwhelmingly invite students to register dissatisfaction: what was unclear, what was too hard, what the instructor did wrong. There are two costs to this framing.

First, it demoralizes staff. When the only signal a lecturer receives is a list of grievances — often delivered anonymously, sometimes cruelly — the rational response is defensiveness or disengagement, not reflective improvement. A deficit instrument teaches teachers to fear feedback rather than seek it.

Second, deficit framing narrows what students report. If the question is "what was the problem?", students supply problems, and the institution never learns what it should protect and amplify. The pedagogically excellent seminar that transformed a struggling student's confidence is invisible to an instrument that only asks what went wrong.

Appreciative Inquiry reframes the core question from "what failed?" to "when did this course work best for you, and how do we do more of that?" This is not naïve positivity; it is a different sampling strategy. Peak-experience stories are unusually rich data — they name the specific practices, moments, and conditions under which learning genuinely happened, which is precisely the information a programme director needs to redesign a course intentionally. Because appreciative questions are experienced as respectful rather than accusatory, they tend to raise engagement and response quality and to make closing the loop more natural: a strength named by students and echoed in a redesign is a visible, motivating win for both students and staff, whereas a fixed complaint is merely a debt repaid.

Critically, the practical case for Appreciative Inquiry in course evaluation is not that it should replace diagnostic evaluation. It is that a mature evaluation system should be able to ask both kinds of question — appreciative and diagnostic — and hold the results side by side.

Limitations and honest caveats

A PhD-level reader will and should raise objections. We take them seriously.

Positivity bias and the risk of ignoring real problems. The most serious critique is structural: an instrument that asks only "when did this work best?" will systematically under-detect failure. A course can generate genuine peak moments for some students while failing others badly. If Appreciative Inquiry is used as the sole evaluation lens, safeguarding concerns, discriminatory conduct, chronic assessment problems, or accessibility failures may never surface because no one was asked. Appreciative Inquiry is emphatically not a substitute for detecting serious failures; any responsible deployment must retain a robust channel for raising problems, including confidential and escalating ones.

Social-desirability and framing effects. Appreciative prompts can cue students toward agreeable answers, compounding the social-desirability pressures already present in evaluation. The generative gains Bushe describes are real, but the same framing that unlocks aspiration can also suppress legitimate dissent if it is not carefully balanced.

The evidence base skews practitioner and case-study. Much of the AI literature — including several sources cited here — is built on case studies, action-research accounts, and integrative reviews rather than randomized controlled trials. This is not disqualifying (the object of study resists RCT designs), but it means claims about effect sizes should be held loosely. Bushe and Kassam's own finding — that only 35% of cases were transformational — is a useful corrective to promotional accounts.

Facilitation dependence. Appreciative Inquiry is highly sensitive to how it is run. A skilled facilitator produces generativity; an unskilled one produces a feel-good session that changes nothing, or worse, papers over problems. Embedding the approach in software does not remove this dependence — it relocates it into question design and analysis.

How Koji incorporates this

Koji for Education is an AI-native course-evaluation platform — and here "AI" means artificial intelligence: the moderator technology that runs conversational interviews. Koji's design lets institutions apply the Appreciative Inquiry philosophy without inheriting its positivity blind spot.

• AI-moderated conversational interviews replace flat form-filling with an adaptive conversation. This is the natural home for appreciative prompts. An institution can frame an interview with strengths-based openers — "Describe a moment this course worked really well for you" or "Tell me about a time you felt genuinely engaged in this module" — and the moderator can probe the specifics of that peak experience, generating exactly the rich, generative data Discovery is meant to surface. Because the moderator is conversational, it can follow an appreciative story into concrete detail rather than stopping at a rating.
• Structured question types — open_ended, scale, single_choice, multiple_choice, ranking, and yes_no — let evaluators build instruments that hold appreciative and diagnostic items together. A ranking item can ask what to protect; an open-ended item can ask a hard diagnostic question. The methodology is a choice, not a constraint of the tooling.
• Automatic thematic analysis surfaces strengths as well as problems. Rather than counting complaints, the analysis identifies recurring peak-experience themes — the practices students name as working — which is precisely the material needed to move from Discovery to Design.
• Balanced, bias-aware reporting. Koji is deliberately built so that positives and negatives both surface. This is the direct answer to the positivity-bias caveat above: an appreciative framing on the front end does not mean a rosy report on the back end. Serious problems raised in any channel are reported, not smoothed over.
• Formative, mid-cycle collection lets an institution run a lightweight appreciative check partway through a course — capturing what is working while there is still time to do more of it — rather than only running a post-mortem after grades are in.
• Closing-the-loop action tracking turns named "peak moments" into concrete design changes and tracks whether they were enacted. This operationalises the Design and Destiny phases: a strength students identify becomes a tracked commitment, and students can see that their appreciative input changed something.

The key claim is a modest and honest one: Koji supports both appreciative and problem-detecting framings. It gives institutions the generative upside that Cooperrider, Srivastva, and Bushe describe, while retaining the diagnostic channels that the limitations section demands. That balance — not appreciative framing alone — is the point.

Koji's core platform at koji.so applies the same conversational-interview and analysis engine to product and customer research, where the same tension between hearing what delights users and detecting what is broken plays out in a commercial setting.

References

• Bushe, G. R., & Kassam, A. F. (2005). When is appreciative inquiry transformational? A meta-case analysis. Journal of Applied Behavioral Science, 41(2), 161–181. https://doi.org/10.1177/0021886304270337
• Cho, H., & Ardichvili, A. (2024). Appreciative Inquiry: An integrative review of studies in three disciplines. Human Resource Development Review, 23(3), 376–401. https://doi.org/10.1177/15344843241256156
• Cooperrider, D. L., & Srivastva, S. (1987). Appreciative inquiry in organizational life. In R. W. Woodman & W. A. Pasmore (Eds.), Research in Organizational Change and Development (Vol. 1, pp. 129–169). JAI Press.
• Sacco-Bene, C. (2022). Appreciative inquiry: A component of course evaluation and improvement. The Journal of Humanistic Counseling, 61(3), 170–183. https://doi.org/10.1002/johc.12180

Related Resources

• Closing the feedback loop: course evaluation evidence
• Do student evaluations improve teaching? The feedback-intervention view
• Students as partners: co-designing course evaluation
• Mid-semester feedback consultation: a meta-analysis
• Utilization-focused evaluation: designing course feedback for use
• Students are willing but doubt feedback is used (Spencer & Schmelkin)
• Negativity bias when reading open-text course-evaluation comments