Human-in-the-loop assessment with AI: implications for teacher education in Ibero-American universities
Open accessThe DOI matches its Crossref record; 26 of the 26 references it cites were found in a registry or an index; 5 of the 12 sources its claims cite were checked for support.
Abstract
This scoping review examines how artificial intelligence (AI) reshapes assessment in Ibero-American higher education and specifies the teacher-training capacities and ethical safeguards needed for responsible adoption. Guided by PRISMA procedures and an eligibility scheme based on PPCDO (Population–Phenomenon–Context–Design–Outcomes), we searched Scopus and screened records (2015–2025; English/Spanish/Portuguese), yielding 76 peer-reviewed studies.
Overview
CoChat Summary
TL;DR
This scoping review of 76 peer-reviewed studies (2015–2025) examines how AI reshapes assessment in Ibero-American higher education and what teacher-training capacities and ethical safeguards are needed. It proposes a function-by-purpose taxonomy of assessment technologies and five competency clusters, concluding that systems are technically ready but pedagogically under-specified.
Key findings
- From 11,938 initial Scopus records, filtering and sampling reduced the corpus to 76 studies; AI-labeled tools accounted for 8.9% (n = 23) overall, while LMS-based assessment appeared in 17.4% (n = 45) of references, with Moodle in 35.6% (n = 16) of LMS mentions.
- Interactive quizzes/tests represented 44.8% (n = 13) of assessment tools in that subsection, with ≈70% including interactive elements; rubrics/portfolios were 20.7% (n = 6); video resources 26.9% (n = 18); VR/AR 0.8% (n = 2); programming tools 2.7% (n = 7).
- Exploratory correlations linked evaluation platforms with continuous feedback (r = 0.83, p = 0.002), simulators with practical skills (r = 0.85, p = 0.001), augmented reality with immersion (r = 0.81, p = 0.001), virtual platforms with time management (r = 0.78, p = 0.002), and authoring tools with personalization (r = 0.74, p = 0.003).
- The review specifies five teacher-education competency clusters: feedback literacy with AI; rubric/item validation and traceability; data interpretation and fairness; integrity and transparency in AI-involved assessment; and orchestration of platforms plus moderation/double-marking when AI assists scoring.
Methods
- Scoping review guided by PRISMA procedures and a PPCDO eligibility scheme (Population–Phenomenon–Context–Design–Outcomes), using Scopus as the sole primary database and screening peer-reviewed articles/reviews from 2015–2025 in English, Spanish, or Portuguese.
- Screening proceeded from 11,938 records to 4,617 after subject/document filters, then a probabilistic sample of 355, then 106 after title/abstract screening, yielding 76 studies for extraction.
- Synthesis combined qualitative thematic analysis, quantitative descriptors/frequencies, and exploratory Pearson correlations between tool families and reported advantages, with significance set at p < 0.05; AI was coded only when primary studies explicitly reported generative, predictive/recommender, or automated scoring/feedback features.
Limitations
- Stated in the paper: only Scopus was reviewed, so complementary studies in other databases may be missed; because technology research advances rapidly, readers should consider the reviewed timeframe and compare with more recent work.
- Evident from the text: the correlational analysis is bivariate and does not establish causality, and several outcomes (e.g., motivation) are broader than assessment per se.
- Evident from the text: the single-database design and sampling strategy may constrain coverage and generalizability, although the authors report forward-backward citation chasing to address single-database bias.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
5 Valid · 6 Not valid · 6 Not checked · 82 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“AI can scale feedback and continuous assessment only when educators are prepared to audit outputs, document assistance limits, and communicate rationales-requirements tightly coupled to data governance and algorithmic transparency (Wang, 2024).”
What the sources say
- Partly supportedcites Wang, 2024 sourceExperimental — accuracy not yet measured
The paper says AI systems may assess student assignments, and Table 1 recommends auditing AI systems, documenting how they work, notifying stakeholders, and training educators. It does not say AI scales feedback or continuous assessment only if educators themselves audit outputs, document limits on AI assistance, and communicate rationales; its auditing and documentation recommendations are directed at systems and developers rather than framed as those educator duties. (LLM, conf=0.85)
plainstates 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.
- Partly supported
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Qualitatively, the evidence portrays a field still governed by institutional risk management logics rather than pedagogical transformation.”
plain- 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 result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Where human moderation and feedback triage are embedded-e.g., teachers auditing samples, editing AI comments, and closing loops with brief in-class clarifications-participants describe higher perceived usefulness and fairness.”
plainstates an association- 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 result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“qualitative reports from instructors show value in triangulating dashboards with situated knowledge (attendance patterns, clinical placement feedback, students' selfreports), which tempers the risk of over-reacting to noisy signals.”
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.
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The qualitative evidence reframes 'eÿciency' as a redistribution rather than a net reduction of work: time saved in marking can be reallocated to rubric refinement, audit sampling, student conferencing, and governance activities (consent, disclosures, documentation).”
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.
33 more claims
Create a free account to see all 99 claims and to verify the 1 reference still open.
- It may state two separate claims as one. 45
- Probably not a substantial claim. 28
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 major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ 3.1.4 Search strategy · p. 6
“Document type = Article and Review (peer-reviewed journals)”
§ 3.2.1 Identification · p. 6
“selecting exclusively articles as the document type, reducing the number of studies to 4617.”
Citations
Cites
The PRISMA 2020 statement: An updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · BMJ · doi:10.1136/bmj.n71 · Cited by 108,283
Technological Pedagogical Content Knowledge: A Framework for Teacher Knowledge
Mishra, P., & Koehler, M. J. (2006) · Teachers College Record: The Voice of Scholarship in Education · doi:10.1111/j.1467-9620.2006.00684.x · Cited by 12,604
Components of Education 4.0 in 21st Century Skills Frameworks: Systematic Review
González-Pérez, L. I., & Ramírez-Montoya, M. S. (2022) · Sustainability · doi:10.3390/su14031493 · Cited by 752
Factors that influence teachers’ adoption and integration of ICT in teaching/learning process
Lawrence, J. E., & Tar, U. A. (2018) · Educational Media International · doi:10.1080/09523987.2018.1439712 · Cited by 528
Oke, A., & Fernandes, F. A. P. (2020) · Journal of Open Innovation: Technology, Market, and Complexity · doi:10.3390/joitmc6020031 · Cited by 523 · cited 1×
“The Fourth Industrial Revolution, characterized by the convergence of digital, physical, and biological technologies, is rapidly transforming all sectors of society, including education (Oke and Fernandes, 2020).”
Flores-Vivar, J.-M., & García-Peñalvo, F.-J. (2023) · Comunicar · doi:10.3916/c74-2023-03 · Cited by 306
González-Salamanca, J. C., Agudelo, O. L., & Salinas, J. (2020) · Sustainability · doi:10.3390/su122410366 · Cited by 254
Las TIC y la formación inicial de los docentes. Modelos y competencias digitales
Cabero Almenara, J., & Martínez Gimeno, A. (2019) · Profesorado, Revista de Currículum y Formación del Profesorado · doi:10.30827/profesorado.v23i3.9421 · Cited by 209
Pedagogical Applications of Generative AI in Higher Education: A Systematic Review of the Field
Qian, Y. (2025) · TechTrends · doi:10.1007/s11528-025-01100-1 · Cited by 206 · cited 1×
“Experimental evidence shows that targeted instructional interventions can recalibrate perceived usefulness and ease-of-use of generative AI and reduce over-reliance in assessed tasks (Qian, 2025).”
Bewersdorff, A., Hornberger, M., Nerdel, C., & Schiff, D. S. (2025) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2024.100340 · Cited by 148
Harnessing AI for education 4.0: Drivers of personalized learning.
Barrera Castro, G. P., Chiappe, A., Becerra Rodriguez, D. F., & Sepulveda, F. G. (2024) · Electronic Journal of e-Learning · doi:10.34190/ejel.22.5.3467 · Cited by 132
Self-citationConstruir el conocimiento en la era digital: retos y reflexiones
Marimón, M., Almenara, J. C., Castañeda, L., Salvador, C. C., de Oliveira, J. M., & Rodríguez‐Triana, M. J. (2022) · Revista de Educación a Distancia (RED) · doi:10.6018/red.505661 · Cited by 75
Addressing the use of generative AI in academic writing.
van Niekerk, J., Delport, P. M. J., & Sutherland, I. (2025) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2024.100342 · Cited by 72 · cited 4×
“Experimental evidence shows that when verification and reflection are embedded into assessed tasks, reliance on generative systems becomes more discerning without sacrificing eÿciency, which is precisely the stance needed in teacher training (Van Niekerk et al., 2025).”
InfluentialArtificial Intelligence and Digital Ecosystems in Education: A Review
Rojas, M. P., & Chiappe, A. (2024) · Technology, Knowledge and Learning · doi:10.1007/s10758-024-09732-7 · Cited by 64 · cited 1×
“In parallel, the much-vaunted benefits of AI–automation of low-level grading and personalization–are contingent on the presence of robust digital ecosystems that render algorithms intelligible and data flows governable (Rojas and Chiappe, 2024).”
Self-citationA Framework for Education 4.0 in Digital Education Ecosystems
Chituc, C.-M. (2021) · IFIP advances in information and communication technology · doi:10.1007/978-3-030-85969-5_66 · Cited by 24 · cited 1×
“As a result, the education sector faces a dual challenge: on the one hand, it must adapt its practices to prepare students for an increasingly digital and automated world, and on the other, it must leverage these new technologies to improve the teaching-learning process (Chituc, 2021).”
Arkorful, V., Salifu, I., Arthur, F., & Abam Nortey, S. (2024) · Cogent Education · doi:10.1080/2331186x.2024.2326722 · Cited by 23
Student Equity in the Age of AI-Enabled Assessment
Stephenson, B., & Harvey, A. (2022) · Routledge eBooks · doi:10.4324/9781003293101-14 · Cited by 20
University students’ trust in AI: Examining reliance and strategies for critical engagement.
Abuzar, M., Mahmudulhassan, & Muthoifin (2025) · International Journal of Interactive Mobile Technologies (iJIM) · doi:10.3991/ijim.v19i07.52875 · Cited by 20
Algorithmic decisions in education governance: Implications and challenges.
Wang, Y. (2024) · Discover Education · doi:10.1007/s44217-024-00337-x · Cited by 18 · cited 1×
“AI can scale feedback and continuous assessment only when educators are prepared to audit outputs, document assistance limits, and communicate rationales–requirements tightly coupled to data governance and algorithmic transparency (Wang, 2024).”
AI-Powered Tools for Personalized Learning in Educational Technology
Amoako, K., Asante, A., & Owusu, K. (2024) · International Journal of Technology and Modeling · doi:10.63876/ijtm.v3i1.115 · Cited by 18 · cited 1×
“In practice, this entails cultivating algorithmic literacy, data stewardship, and pedagogical judgment for task redesign toward authentic evidence, competencies repeatedly flagged by high-impact reviews of AI-enabled personalization and platforms (Amoako et al., 2024).”
Cisneros-Barahona, A. S., Marqués-Molías, L., Samaniego-Erazo, G. N., & Mejía-Granizo, C. M. (2024) · RIED-Revista Iberoamericana de Educación a Distancia · doi:10.5944/ried.27.2.39122 · Cited by 12
P. Miranda, J. P., B. Dianelo, R. F., B. Gamboa, A., Bansil, J. A., R. Regala, A., & S. Simpao, L. (2024) · International Journal of Evaluation and Research in Education (IJERE) · doi:10.11591/ijere.v13i5.28919 · Cited by 5
AI in assessment analysis and improvement
Trajkovski, G., & Hayes, H. (2025) · Digital Education and Learning · doi:10.1007/978-3-031-88252-4_4 · Cited by 4
ICT integration in FLT: An analysis of TPACK implementation in Spanish primary teacher education.
Gómez Sánchez, T. F., Bobadilla-Pérez, M., Rumbo Arcas, B., Fraga-Viñas, L., & Galán-Rodríguez, N. M. (2024) · Digital Education Review · doi:10.1344/der.2024.45.214-221 · Cited by 4
Martín-Párraga (2024) · doi:10.24140/issn.1645-7250.rle61.05 · Cited by 4
International Journal of Technology and Modeling
(2025) · International Journal of Technology and Modeling · doi:10.63876/ijtm
Research works printed but not linked: 1
- Chaparro-Martínez, 2016, Gestión de la información: Uso de las bases de datos scopus y web of science con fines académicos., Univers. Ciencia Tecnol., 20, 166
Cited by
Lazaros, K., Vrahatis, A. G., & Kotsiantis, S. (2026) · Entropy · doi:10.3390/e28040377 · Cited by 50
Artificial intelligence in doctoral training: Bridging gaps, easing burdens
Rodríguez, G. A., Chiappe, A., & Durand, J. C. (2026) · Social Sciences & Humanities Open · doi:10.1016/j.ssaho.2026.102744 · Cited by 5
Self-citationDehghan, F. (2026) · Cogent Education · doi:10.1080/2331186x.2026.2613501 · Cited by 3
Smahi, K., Labouidya, O., & Khadiri, K. E. (2026) · E3S Web of Conferences · doi:10.1051/e3sconf/202669801010 · Cited by 1
“Georgieva & Stuart emphasize that any deployment of AI in higher education should be based on a clear ethical framework, with principles of transparency, accountability, and respect for learner autonomy [9].”
From Learning-Style Neuromyths to AI-Enabled Personalisation for Teacher Preparation
Casallas, L. C. E., Chiappe, A., & Sáez-Delgado, F. (2026) · Improving Schools · doi:10.1177/13654802261432037 · Cited by 6
“…focus, we determined study inclusion using the PPCDO framework—Population, Phenomenon, Context, Design, and Outcomes—which has recently been adopted in AI-and teacher-education reviews to align eligibility decisions with downstream synthesis and training implications (Fajardo-Ramos et al., 2025).”
Self-citation
Metadata
- Authors
- Fajardo-Ramos, D. C., Chiappe, A., & Mella-Norambuena, J.
- Affiliations
- Universidad de La Sabana (CO)
- University of the Americas (CL)
- Corresponding author
- Chiappe, Andrés
- Venue
- Frontiers in Education
- Publisher
- Frontiers Media SA
- Published
- 2025-10-23
- Volume
- 10
- ISSN
- 2504-284X (ISSN-L)
- OpenAlex
- W4415457454
- Cited by
- 29
- References
- 24
- Citations by year
- 2026: 29
- FWCI
- 30.15
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Frontiers in Education (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Universidad de La Sabana
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Social Sciences › Education › Student Assessment and Feedback
- Social Sciences › Education › Educational Technology and Pedagogy
- Keywords
- AI-assisted assessment, teacher professional development, Ibero-American higher education, AI-assisted grading, generative AI, adaptive testing, algorithmic feedback, explainability, data stewardship, audit trail, rubric validation, duplicate marking, AI ethics in education
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 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (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
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.