The DOI matches its Crossref record; 17 of the 17 references it cites were found in a registry or an index; 2 of the 33 sources its claims cite were checked for support.
From Cognitive Necessity to Cognitive Choice: Higher Education Assessment and Learning in the Age of Generative AI
Open accessAbstract
The widespread adoption of generative artificial intelligence in higher education has intensified debates around assessment, authorship, and academic integrity. This paper argues that such debates obscure a more fundamental pedagogical shift, namely, the decoupling of assessment performance from cognitive engagement. Historically, assessment functioned not only as a measure of learning, but also as a structural mechanism that implicitly enforced cognitive engagement. With the advent of GenAI, learners can increasingly produce assessment outputs without necessarily engaging in the cognitive processes traditionally associated with learning. As a result, cognitive engagement has shifted from being a pedagogical necessity to an intentional learner choice. This paper conceptualises this shift as the cognitive engagement gap, wherein successful assessment completion no longer reliably indicates learning or epistemic development. Through a theory-informed conceptual analysis, the paper examines how GenAI reconfigures learning processes, challenges the validity of assessment as a proxy for learning, and exposes long-standing assumptions embedded in assessment-centred pedagogies. In response, the paper proposes a Cognitive Engagement-Centred Assessment (CECA) framework, offering principled guidance for designing assessment that foregrounds cognitive processes, metacognition, and learning assurance in AI-mediated environments. The paper concludes by positioning GenAI not as a threat to assessment, but as a catalyst for more intentional, transparent, and learning-centred pedagogical design.
Overview
CoChat Summary
TL;DR
This conceptual paper argues that generative AI has shifted cognitive engagement in higher education from a structural necessity enforced by assessment design to an intentional learner choice, producing a "cognitive engagement gap" in which successful assessment performance no longer reliably indicates learning. It proposes the Cognitive Engagement-Centred Assessment (CECA) framework as a design-oriented corrective.
Key findings
No empirical findings are reported; the paper is a theory-informed conceptual analysis and explicitly states that references to "effects," "implications," or "outcomes" are "theoretical and analytical interpretations derived from prior scholarship, not as empirically observed results from the present study."
- The paper's central construct is the *cognitive engagement gap*: "the growing separation between assessment performance and the cognitive processes that underpin meaningful learning," meaning assessment completion "no longer implies meaningful cognitive engagement or learning."
- It argues GenAI is a "structural catalyst rather than a root cause," amplifying pre-existing vulnerabilities in assessment-centred pedagogy (surface learning, strategic compliance) rather than creating them.
- It identifies equity risks: when engagement becomes optional, outcomes depend more on learners' prior preparation, academic confidence, and self-regulatory capacity, so "generative AI risks amplifying existing educational inequalities."
- It proposes CECA with five principles: cognitive explicitness, process visibility, AI-agnostic robustness, metacognitive accountability, and pedagogical assurance over surveillance.
Methods
Design: theory-informed conceptual analysis, not empirical research; the paper states it does not report empirical data or test hypotheses.
- Approach: systematic analysis and synthesis of literature in assessment theory, learning sciences, and AI-in-education, proceeding through "iterative processes of theoretical comparison, abstraction, and construct development."
- Output: three interrelated conceptual constructs — cognitive necessity, cognitive choice, and the cognitive engagement gap — "advanced as analytical lenses rather than empirical claims," plus the CECA framework.
Limitations
Stated by the paper: "As a conceptual contribution, this paper does not provide empirical validation of the cognitive engagement gap or the CECA framework"; it calls for future research on how assessment designs mediate cognitive engagement, learner experiences of cognitive choice, disciplinary differences in vulnerability, and effects of CECA-aligned designs on learning outcomes, equity, and student agency.
- Stated by the paper: the analysis draws on prior scholarship rather than new data, so its claims are interpretive.
- Evident from the text: the CECA framework is presented as "not offered as a prescriptive model" and its operational best practices are cited to the author's own prior conference papers (Montebello, 2025a, 2025b), which are not detailed in this article; the paper also notes that alternative formats such as oral and dialogic assessment "raise legitimate concerns around scalability and workload" without resolving them.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
2 Valid · 1 Not valid · 20 Not checked · 72 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Generative AI intensifies these demands by offering high-quality performance substitutes that can mask gaps in understanding.”
What the sources say
- Not checked yetcites Zimmerman, 2002 source
- Partly supportedcites Nicol & Macfarlane-Dick, 2006 sourceExperimental — accuracy not yet measured
The conceptual model (p. 4) distinguishes internal outcomes such as increased understanding from externally observable products such as essays, and the Introduction (p. 2) describes students monitoring their learning through self-regulation. The paper does not address generative AI or say that high-quality substitute work can mask gaps in understanding or intensify demands on self-regulation. (LLM, conf=0.82)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“At the same time, the reproduction of biases and representational patterns embedded in training data raises concerns about whose knowledge, styles, and voices are normalised in AI-mediated learning contexts (Bender et al., 2021; Tao et al., 2024; Sterlie et al., 2024).”
What the sources say
- Not checked yetcites Bender et al., 2021
- Supportedcites Tao et al., 2024 source
The Introduction (page unavailable) links overrepresentation in LLM training corpora to cultural bias in generated language and notes that LLMs are used in education. The Discussion (page unavailable) reports that model outputs favor Western cultural values and raises concerns about cultural misrepresentation and effects on users’ expression. The citing sentence frames normalization in learning contexts as a concern, not as an observed educational outcome. (LLM, conf=0.86)
- Not checked yetcites Sterlie et al., 2024
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“Policy analyses further indicate that many university guidelines frame generative AI primarily through the lenses of originality and misconduct, rather than through a reconsideration of what assessment is intended to evidence (Luo, 2024).”
What the sources say
- Not checked yetcites Luo, 2024
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“When policy attention focuses narrowly on unauthorised use, it risks preserving the assumption that traditional assessment outputs remain stable indicators of learning, an assumption this paper challenges.”
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“The emergence of GenAI disrupts this long-standing relationship between assessment and learning by introducing a structural condition in which acceptable academic artefacts can be produced without the learner necessarily engaging in the cognitive processes traditionally associated with learning (Cope et al., 2020; Dawson et al., 2023).”
- Also stated in: § 2. Assessment and Learning Before GenAI · plain
“GenAI introduces the first scalable condition in which acceptable academic performance can occur without such enactment.”
What the sources say
- Not checked yetcites Cope et al., 2020 source
- Not checked yetcites Dawson et al., 2023
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
- Also stated in: § 2. Assessment and Learning Before GenAI · plain
55 more claims
Create a free account to see all 95 claims and to verify the 31 references still open.
- Probably not a substantial claim. 20
- It may state two separate claims as one. 18
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ References
“Data Availability Statement: The raw data supporting the conclusions of this article will be made available by the authors on request.”
§ 1. Introduction
“Rather than reporting empirical data or testing hypotheses, the paper systematically analyses and synthesises the established bodies of literature in assessment theory, learning sciences, and AI-in-education”
Citations
Cites
Becoming a Self-Regulated Learner: An Overview
Zimmerman, B. J. (2002) · Theory Into Practice · doi:10.1207/s15430421tip4102_2 · Cited by 10,197 · cited 6×
“Research on self-regulated learning consistently demonstrates that learners vary substantially in their ability to plan, monitor, and evaluate their own understanding, particularly in complex academic tasks (Zimmerman, 2002; Nicol & Macfarlane-Dick, 2006).”
InfluentialOn the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988 · cited 1×
“At the same time, the reproduction of biases and representational patterns embedded in training data raises concerns about whose knowledge, styles, and voices are normalised in AI-mediated learning contexts (Bender et al., 2021; Tao et al., 2024; Sterlie et al., 2024).”
Nicol, D. J., & Macfarlane‐Dick, D. (2006) · Studies in Higher Education · doi:10.1080/03075070600572090 · Cited by 5,846
Formative assessment and the design of instructional systems
Sadler, D. R. (1989) · Instructional Science · doi:10.1007/bf00117714 · Cited by 4,285 · cited 7×
“…have been shown to adopt performance-oriented and efficiency-driven strategies, particularly under conditions of high workload and assessment pressure, pri-oritising task completion over conceptual understanding (Sadler, 1989; Zimmerman, 2002; Nicol & Macfarlane-Dick, 2006; Boud & Falchikov, 2007).”
InfluentialChatting and cheating: Ensuring academic integrity in the era of ChatGPT
Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2024) · Innovations in Education and Teaching International · doi:10.1080/14703297.2023.2190148 · Cited by 3,087 · cited 2×
“Coherent, discipline-appropriate texts can now be generated that satisfy many surface and mid-level assessment criteria, including structure, argumentation, and academic register (Cotton et al., 2023; Luo, 2024).”
Teacher Technology Change: How Knowledge, Confidence, Beliefs, and Culture Intersect
Ertmer, P. A., & Ottenbreit-Leftwich, A. T. (2010) · Journal of Research on Technology in Education · doi:10.1080/15391523.2010.10782551 · Cited by 2,556
Rethinking Assessment in Higher Education: Learning for the Longer Term
(2007) · doi:10.4324/9780203964309 · Cited by 1,049
Historical threads, missing links, and future directions in AI in education
Williamson, B., & Eynon, R. (2020) · Learning, Media and Technology · doi:10.1080/17439884.2020.1798995 · Cited by 981
Kapur, M. (2008) · Cognition and Instruction · doi:10.1080/07370000802212669 · Cited by 867 · cited 4×
“Generative AI amplifies these weaknesses by removing the implicit requirement for cognitive struggle, productive failure, and metacognitive reflection (Kapur, 2008), thereby challenging the long-held assumption that assessment participation naturally enforces learning.”
InfluentialArtificial intelligence for education: Knowledge and its assessment in AI-enabled learning ecologies
Cope, B., Kalantzis, M., & Searsmith, D. (2021) · Educational Philosophy and Theory · doi:10.1080/00131857.2020.1728732 · Cited by 518 · cited 3×
“From a sociotechnical and learning-ecologies perspective, generative AI becomes part of a broader assemblage of tools and practices that co-produce learning experiences (Cope & Kalantzis, 2017; Cope et al., 2020).”
Research works printed but not linked: 9
- Biggs, J., and Tang, C. (2011). Teaching for quality learning at university, OU Press. [4th ed.].
- Black, 1998, Assessment and classroom learning, Assessment in Education: Principles, Policy & Practice, 5, 7
- Dawson, 2023, Artificial intelligence and the future of education: Teaching and learning in the age of generative AI, Computers & Education: Artificial Intelligence, 4, 100124
- (EU) 2024/1689 laying down harmonised rules on artificial intelligence (artificial intelligence Act). Official journal of the European Union, L1689, European Parliament & Council of the European Union.
- Laurillard, D. (2012). Teaching as a design science: Building pedagogical patterns for learning and technology, Routledge.
- Luo, 2024, Academic integrity in the age of generative AI: Policy responses and pedagogical implications, AI in Education, 1, 5
- Montebello, M. (, January August). Educating the educators: Guidelines for ethical and empowering use of generative ai in higher education. 9th International Conference on Modern Research in Education, Teaching, and Learning (ICMETL), Oxford, UK.
- Montebello, M. (, January November). Generative AI and the human research scholar. 18th International Conference on Education, Research and Innovation (ICERI25), Seville, Spain.
- Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes, Harvard University Press.
Cited by
Giordmaina, J., & Zammit, L. (2026) · Frontiers in Education · doi:10.3389/feduc.2026.1996379
Metadata
- Authors
- Montebello, M.
- Affiliations
- University of Malta (MT)
- Corresponding author
- Montebello, Matthew
- ORCID
- Montebello, Matthew
- Venue
- AI in Education
- Publisher
- MDPI AG
- Published
- 2026-04-16
- Volume
- 2
- Issue
- 2
- Pages
- 12
- ISSN
- 3042-8130 (ISSN-L)
- OpenAlex
- W7154580082
- Cited by
- 1
- References
- 17
- Citations by year
- 2026: 1
- FWCI
- 8.51
- Type
- journal article
- Language
- en
- Open access
- diamond
- doi.org
- Licence
- Locations
- AI in Education (publishedVersion, cc-by)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Social Sciences › Education › Student Assessment and Feedback
- Social Sciences › Education › Academic integrity and plagiarism
- Keywords
- generative AI, higher education assessment, cognitive engagement, assessment validity, metacognition, assurance of learning, AI-mediated learning, assessment design
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: From Cognitive Necessity to Cognitive Choice: Higher Education Assessment and Learning in the Age of Generative AI. https://app.cochat.ai/papers/10.3390/aieduc2020012
- The paper
- Montebello, M. (2026). From Cognitive Necessity to Cognitive Choice: Higher Education Assessment and Learning in the Age of Generative AI. AI in Education. https://doi.org/10.3390/aieduc2020012
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.