Artificial Intelligence for Student Assessment: A Systematic Review
Open accessThe DOI matches its Crossref record; 60 of the 61 references it cites were found in a registry or an index; 19 of the 73 sources its claims cite were checked for support.
Abstract
This paper analyzes the use of AI for student assessment based on a systematic review and concludes that pedagogy is often not reflected, formative evaluation is common, and more educationally focused research is needed.
Overview
CoChat Summary
TL;DR
This systematic review of 22 studies, selected from 454 initial records, finds that AI in student assessment is used mainly for formative assessment and automatic grading, while pedagogical models are often not reflected. It concludes that teacher training and more pedagogically focused research are needed, especially beyond higher education.
Key findings
- From 454 papers found, 449 remained after duplicates were removed, and 22 were selected; most selected studies focused on higher education, with some examples at secondary level, and the predominant author country was the USA (4 contributions), followed by Greece (3).
- Of the 22 selected papers, 15 had AI for formative assessment as their starting point, and 10 explicitly stated that the system used grades students automatically.
- Only one study clearly addressed pedagogical models underpinning AI use; three others mentioned pedagogical rationale, mainly collaboration or argument maps, and recent work focused more on technical aspects than pedagogy.
- Three studies compared AI use with non-use: Goel and Joyner found students who used AI performed better; Grivokostopoulou et al. found AI and manual grading were very similar and correlated; Samarakou et al. found those who used AI scored 17% higher than those who did not.
Methods
- Systematic review following the PRISMA Statement, searching Scopus and Web of Science in March 2021 with terms including “Artificial Intelligence,” “Machine Learning,” “Education,” “Assessment,” and “Evaluation”; inclusion criteria were peer-reviewed empirical research published 2010–2020 in English or Spanish.
- Titles and abstracts were evaluated independently by two reviewers using Rayyan in blind mode, with a third party resolving discrepancies; full texts were then reviewed, and coding/data extraction used Rayyan tags, a spreadsheet, and Mendeley.
- After screening, 339 studies were excluded after title/abstract review, 68 full texts were discarded for not meeting criteria, and 10 were excluded for lack of access, leaving 22 studies for analysis.
Limitations
- The paper states that the study found a small number of articles focused on student assessment relative to the large number of AI applications, and it calls for further research, mainly at educational levels other than higher education, plus teacher training.
- The paper states that the method was applied rigorously to guarantee validity relative to the eligibility criteria.
- Evident from the text: the search was limited to Scopus and Web of Science, English and Spanish, and 2010–2020; 10 full texts were excluded because they were not accessible; the final set of 22 studies was mostly higher education and STEM/health contexts, limiting generalizability.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
16 Valid · 4 Not valid · 24 Not checked · 43 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“three of them determine the need to give early feedback as an essential aspect in the educational process [22,37,42]”
§ Artificial Intelligence for Student Assessment: A Systematic Review > 4.2. Pedagogical Model Used
What the sources say
- Not matchedcites [22] Samarakou, 2016, Evaluation of an intelligent open learning system for engineering education, Knowl. Manag. E-Learning Int. J.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Not matchedcites [37] Ouguengay, 2015, A neuro-fuzzy inference system for the evaluation of reading/writing competencies acquisition in an e-learning environnement, J. Theor. Appl. Inf. Technol.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Supportedcites [42] Grivokostopoulou, 2017, An Educational System for Learning Search Algorithms and Automatically Assessing Student Performance, Int. J. Artif. Intell. Educ. source
The Feedback Provision section (page not provided) calls feedback an important factor in learning and says the system provides meaningful, immediate feedback. This supports the claim as it applies to this cited work; the other two cited works are not assessed here. (LLM, conf=0.91)
plain- 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.
- Rests on an assumption not shown here.
- Not matched
Every source the paper cites for this claim was checked and supports it.
In the paper
“There is only one study [18] that clearly addresses aspects related to pedagogical models that support the work to be developed subsequently.”
§ Artificial Intelligence for Student Assessment: A Systematic Review > 4.2. Pedagogical Model Used
What the sources say
- Supportedcites [18] Conejo, 2013, Statistical Techniques to Explore the Quality of Constraints in Constraint-Based Modeling Environments, Int. J. Artif. Intell. Educ. source
The paper is a study of constraint-based models for learning environments. Its Constraint-Based Modeling section says that estimates of student knowledge drive instructional strategies and discusses improving a tutor’s pedagogical actions. The claim’s judgment that this is the only relevant study belongs to the citing authors, not to the cited paper. (LLM, conf=0.84)
boostedSingle case- 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.
- Rests on an assumption not shown here.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“On the other hand, the work of Grivokostopoulou et al. [42] made a comparison between the results obtained with AI and the teaching assessments made by hand to check the accuracy of this technology.”
§ Artificial Intelligence for Student Assessment: A Systematic Review > 4.5. Comparison between IA Use and Non-Use
What the sources say
- Supportedcites [42] Grivokostopoulou, 2017, An Educational System for Learning Search Algorithms and Automatically Assessing Student Performance, Int. J. Artif. Intell. Educ. source
The Results section says the authors compared the automated assessment system with the tutor’s assessment, using accuracy and other metrics. The passage does not give a page number. (LLM, conf=0.94)
plain- 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.
- Supported
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
“Three other studies [21,38,39] also hint at their pedagogical rationale, which is mainly based on collaboration.”
§ Artificial Intelligence for Student Assessment: A Systematic Review > 5. Discussion and Conclusions
What the sources say
- Supportedcites [21] Santos, 2014, Involving Users to Improve the Collaborative Logical Framework, Sci. World J. source
The Background describes collaborative learning as engaging students through social interaction and developing teamwork skills. That supports the citing sentence’s characterization of this study’s pedagogical rationale as mainly based on collaboration. (LLM, conf=0.91)
- Supportedcites [38] Kaila, 2016, Redesigning an Object-Oriented Programming Course, ACM Trans. Comput. Educ. source
The Abstract (page unavailable) says the redesign focused on active learning methods and student collaboration. The Introduction and “Step 2: Encouraging Collaboration” (pages unavailable) explain why collaboration was incorporated and how it shaped the tutorials. This supports the citing sentence’s hedged characterization of the study’s pedagogical rationale. (LLM, conf=0.88)
- Not checked — no full text of the sourcecites [39] Rapanta, 2016, The Use of Argument Maps as an Assessment Tool in Higher Education, Int. J. Educ. Res. source
hedgedstates 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.
- Supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“As for the papers that focus on formative assessment, it is clear that the idea of using AI for this task lies in the help that this technology can provide to teachers when they have a large number of students [41], or based on the idea that learning improves the more immediate the feedback given to students [21].”
§ Artificial Intelligence for Student Assessment: A Systematic Review > 5. Discussion and Conclusions
What the sources say
- Could not verifycites [21] Santos, 2014, Involving Users to Improve the Collaborative Logical Framework, Sci. World J. sourceExperimental — accuracy not yet measured
Section 4, “Formative Evaluation” (page unavailable), describes user studies to improve the design of a collaborative learning tool, not formative assessment of students. Section 3.3.3 describes personalized recommendations, but does not say that learning improves when students receive feedback sooner. Searches for immediate feedback and timely feedback found no passage establishing the attributed claim. (LLM, conf=0.76)
boostedstates 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.
- Could not verify
36 more claims · 2 with their cited sources checked
Create a free account to see all 87 claims and to verify the 3 references still open.
- Probably not a substantial claim. 30
- It may state two separate claims as one. 16
- Borderline: one of the tests for a substantial claim is unclear. 1
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“It can revolutionize the way assessments are conducted by providing real-time feedback to students [23].”
From New Era of Artificial Intelligence in Education: Towards a Sustainable Multifaceted Revolution (2023) · cites this paper as [23]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Specifically, AI can be trained to grade a range of subjects and assessments including essays, problem-solving questions, and even graphing assessments [23].”
From New Era of Artificial Intelligence in Education: Towards a Sustainable Multifaceted Revolution (2023) · cites this paper as [23]
Experimental — accuracy not yet measuredIn the citing paper
“These outcomes highlight the fact that machines are far from human intelligence and emotions and cannot undertake the role of teachers (González-Calatayud et al., 2021;Zhang & Aslan, 2021).”
From A systematic review of literature reviews on artificial intelligence in education (AIED): a roadmap to a future research agenda (2024) · cites this paper as González-Calatayud et al., 2021
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
§ Artificial Intelligence for Student Assessment: A Systematic Review > 5. Discussion and Conclusions
“| Phattanapon et al. [36] | 2011 | Artificial intelligence in medicine | Thailand | Computer science and Dentistry | Dental study |”
§ Artificial Intelligence for Student Assessment: A Systematic Review > References
“36. Rhienmora, P.; Haddawy, P.; Suebnukarn, S.; Dailey, M.N. Intelligent dental training simulator with objective skill assessment and feedback.”
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
§ Artificial Intelligence for Student Assessment: A Systematic Review > 5. Discussion and Conclusions
“only 22 studies are mainly focused on student assessment.The majority of them have been published in the last five years, and in the USA.”
§ Artificial Intelligence for Student Assessment: A Systematic Review > 5. Discussion and Conclusions
“| USA | 4 |”
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
“Of the 449 articles reviewed after eliminating duplicates, a total of 339 studies were excluded because they did not comply with the established criteria after reading the abstract and title.The full text of the remaining 100 papers was reviewed”
§ Artificial Intelligence for Student Assessment: A Systematic Review > 4. Results
Reported 100 · recomputed 110
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
§ Artificial Intelligence for Student Assessment: A Systematic Review > 4.5. Comparison between IA Use and Non-Use
“four levels related to student grades”
§ Artificial Intelligence for Student Assessment: A Systematic Review > 4.5. Comparison between IA Use and Non-Use
“very low, low, medium, good, and excellent”
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
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J., Bossuyt, P., Boutron, I., Hoffmann, T., mulrow, C., Shamseer, L., Tetzlaff, J., Akl, E., Brennan, S. E., et al. (2020) · doi:10.31222/osf.io/v7gm2_v1 · Cited by 8,090
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J., Bossuyt, P., Boutron, I., Hoffmann, T., mulrow, C., Shamseer, L., Tetzlaff, J., Akl, E., Brennan, S. E., et al. (2020) · doi:10.31222/osf.io/v7gm2 · Cited by 5,177
An Introduction to Systematic Reviews
Gough, D., Oliver, S., & Thomas, J. (2017) · doi:10.4135/9781036234942 · Cited by 1,239
Chatterjee, S., & Bhattacharjee, K. K. (2020) · Education and Information Technologies · doi:10.1007/s10639-020-10159-7 · Cited by 747 · cited 1×
“Chatterjee and Bhattacharjee [26] analyze AI to manage resources on educational organizations.”
BackgroundHuman- versus Artificial Intelligence
Korteling, J. E. (., van de Boer-Visschedijk, G. C., Blankendaal, R. A. M., Boonekamp, R. C., & Eikelboom, A. R. (2021) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2021.622364 · Cited by 619
Artificial 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×
“The comparison between human and artificial intelligence is really interesting [7,8].”
BackgroundDesign of online intelligent English teaching platform based on artificial intelligence techniques
Sun, Z., Anbarasan, M., & Praveen Kumar, D. (2021) · Computational Intelligence · doi:10.1111/coin.12351 · Cited by 409 · cited 3×
“[46] indicates that “artificial intelligence (AI) is the combination of intelligence, that is, machines capable of demonstrating human intelligence and of making decisions with human skills”; AI is therefore intended to “build highly advanced machines that can make smart decisions” (p.”
BackgroundEnabling Technologies for Urban Smart Mobility: Recent Trends, Opportunities and Challenges
Paiva, S., Ahad, M., Tripathi, G., Feroz, N., & Casalino, G. (2021) · Sensors · doi:10.3390/s21062143 · Cited by 395 · cited 1×
“AI is applied to different contexts, so we can see applications to architecture in smart buildings and smart cities [6], smart mobility [9], medicine [10–12], smart industry (the 4th industrial revolution, as some authors [13] remark) and also smart education or smart classrooms [14].”
BackgroundInteligencia artificial y sus implicaciones en la educación superior
Ocaña-Fernández, Y., Valenzuela-Fernández, L. A., & Garro-Aburto, L. L. (2019) · Propósitos y Representaciones · doi:10.20511/pyr2019.v7n2.274 · Cited by 358
Research works printed but not linked: 1
- Ouguengay, 2015, A neuro-fuzzy inference system for the evaluation of reading/writing competencies acquisition in an e-learning environnement, J. Theor. Appl. Inf. Technol., 81, 600
Cited by
Chiu, T. K. F., Xia, Q., Zhou, X., Chai, C. S., & Cheng, M. (2023) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2022.100118 · Cited by 1,181
New Era of Artificial Intelligence in Education: Towards a Sustainable Multifaceted Revolution
Kamalov, F., Santandreu Calonge, D., & Gurrib, I. (2023) · Sustainability · doi:10.3390/su151612451 · Cited by 1,089
“Specifically, AI can be trained to grade a range of subjects and assessments including essays, problem-solving questions, and even graphing assessments [27].”
BackgroundArtificial Intelligence in Education: A Systematic Literature Review
Wang, S., Wang, F., Zhu, Z., Wang, J., Tran, T., & Du, Z. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124167 · Cited by 977
Chatbots in Education and Research: A Critical Examination of Ethical Implications and Solutions
Kooli, C. (2023) · Sustainability · doi:10.3390/su15075614 · Cited by 666
Artificial Intelligence in Education: AIEd for Personalised Learning Pathways
Tapalova, O., & Zhiyenbayeva, N. (2022) · Electronic Journal of e-Learning · doi:10.34190/ejel.20.5.2597 · Cited by 633
Metadata
- Authors
- González-Calatayud, V., Prendes-Espinosa, P., & Roig-Vila, R.
- Affiliations
- Department of Statistics, Mathematics and Computer Sciences, Miguel Hernández University, 03202 Elche, Spain
- Department of Didactic and School Organization, University of Murcia, 30100 Murcia, Spain
- University of Alicante (ES)
- Venue
- Applied Sciences
- Publisher
- MDPI AG
- Published
- 2021-06-12
- Volume
- 11
- Issue
- 12
- Pages
- 5467
- ISSN
- 2076-3417 (ISSN-L)
- OpenAlex
- W3171148899
- Semantic Scholar
- 783dead4fd1aa29ee019a2f324e859a6b3dcb332
- Cited by
- 503
- References
- 59
- Citations by year
- 2026: 87 · 2025: 167 · 2024: 169 · 2023: 51 · 2022: 24 · 2021: 5
- FWCI
- 9.46
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Applied Sciences (publishedVersion, cc-by)
- Repositorio Institucional de la Universidad de Alicante (Universidad de Alicante) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
- Social Sciences › Education › Educational Technology and Pedagogy
- Keywords
- artificial intelligence in education, student assessment, automated grading, formative assessment, AI-based tutoring, teacher professional development
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Artificial Intelligence for Student Assessment: A Systematic Review. https://app.cochat.ai/papers/10.3390/app11125467
- The paper
- González-Calatayud, V., Prendes-Espinosa, P., & Roig-Vila, R. (2021). Artificial Intelligence for Student Assessment: A Systematic Review. Applied Sciences. https://doi.org/10.3390/app11125467
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.