The DOI matches its Crossref record; 71 of the 71 references it cites were found in a registry or an index; none of the 9 sources its claims cite is checked for support yet.
ChatGPT for Education and Research: Opportunities, Threats, and Strategies
Open accessPublished version of a preprint
Abstract
In recent years, the rise of advanced artificial intelligence technologies has had a profound impact on many fields, including education and research. One such technology is ChatGPT, a powerful large language model developed by OpenAI. This technology offers exciting opportunities for students and educators, including personalized feedback, increased accessibility, interactive conversations, lesson preparation, evaluation, and new ways to teach complex concepts. However, ChatGPT poses different threats to the traditional education and research system, including the possibility of cheating on online exams, human-like text generation, diminished critical thinking skills, and difficulties in evaluating information generated by ChatGPT. This study explores the potential opportunities and threats that ChatGPT poses to overall education from the perspective of students and educators. Furthermore, for programming learning, we explore how ChatGPT helps students improve their programming skills. To demonstrate this, we conducted different coding-related experiments with ChatGPT, including code generation from problem descriptions, pseudocode generation of algorithms from texts, and code correction. The generated codes are validated with an online judge system to evaluate their accuracy. In addition, we conducted several surveys with students and teachers to find out how ChatGPT supports programming learning and teaching. Finally, we present the survey results and analysis.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
This paper examines ChatGPT’s opportunities and threats for education and research, with a focus on programming learning and teaching. It combines coding experiments on an online judge system with surveys of students and teachers, finding generally useful support but also accuracy, integrity, and over-reliance concerns.
Key findings
- In code-generation experiments, ChatGPT’s generated code was correct about 95.83% of the time on a basic compiler and about 75% on the Aizu Online Judge (AOJ), with an average accuracy of 85.42%.
- Among surveyed students, 78.8% had used ChatGPT for programming problems, 86.7% said its suggestions helped solve programming problems, and 92.9% found it useful for learning programming; most student satisfaction ratings were 4/5 (36.6%) or 5/5 (20%).
- About 60% of surveyed teachers had used ChatGPT for teaching programming, but about 50% rated satisfaction as 3/5, suggesting limited trust; one teacher said ChatGPT “is still not perfect to answer some exams of my Java class.”
- The paper identifies threats including cheating on online exams and assignments, diminished critical thinking, difficulty evaluating AI-generated text/code, ethical implications and bias, and blind reliance on generative AI; proposed strategies include clear assignment instructions, plagiarism detection, proctoring, comparing responses, and follow-up questions.
Methods
- The authors conducted coding-related experiments with ChatGPT: code generation from problem descriptions, pseudocode generation, code correction/debugging, and code optimization.
- For code generation, they used eight random “Algorithms and Data Structures” problems from AOJ, generated code three times per problem, and validated results on both a basic compiler and the AOJ platform.
- They also conducted surveys with undergraduate, Master’s, and doctoral students and with teachers about ChatGPT’s support for programming learning and teaching; the paper reports percentages but not the total number of respondents.
Limitations
- The paper states that experiments were conducted on ChatGPT during active development, so results may vary with new ChatGPT versions, different questions, different problem descriptions, and different solution codes for optimization.
- The paper notes that although ChatGPT generated about 85.42% correct code on average, adjustments may still be necessary for error-free compilation and acceptance, and it failed an elementary counting task involving the digit “8.”
- Evident from the text: the survey sample size is not reported, and the coding experiments are limited to eight AOJ problems with three generations each, which restricts generalizability.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
38 Own interpretation · 23 Own result · 6 Not checked yet · 5 No basis given · 1 Own method
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“the correctness rate of the generated code based on the basic compiler is approximately 95.83%”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.2. Solution Code Generation
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 reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“ChatGPT generates mostly correct code considering long problem descriptions, including constraints, algorithms, and input and output formatting.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.2. Solution Code Generation
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 reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“the correctness rate based on AOJ compilation is approximately 75%”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.2. Solution Code Generation
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.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In addition, we experimented with the ChatGPT model to find the derivatives of mathematical equations, and it solved all equations correctly.In this case, the correctness is 100%.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 3.1. Opportunities for Learners
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.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We also asked ChatGPT about quizzes of other levels, such as medium and easy, and we found that there was a significant difference between quizzes of different levels.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Rapid Assessment and Evaluation:
plaincorrelational- 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.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“ChatGPT can help optimize codes by suggesting ways to reduce the memory usage and time complexity.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.4. Solution Code Optimization
plainconditionalSingle case- 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 reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, we asked ChatGPT to “prepare a detailed outline for the Algorithms and Data Structures course”, as depicted in Figure 2. It creates a complete table of contents for the Algorithms and Data Structures course, showing topics and a breakdown of each topic, along with learning objectives.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Lesson Planning:
plaindirectSingle 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.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We found that most students rated their programming experiences between 1/5 and 3/5.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 5. Survey on ChatGPT Language Model for Programming Learning and Teaching Support
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.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The average accuracy of the code generated by ChatGPT is 85.42%.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 7. Limitations
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 reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We found that about 86.7% of the students answered “Yes”.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 5. Survey on ChatGPT Language Model for Programming Learning and Teaching Support
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.
- It may state two separate claims as one. 18
- Probably not a substantial claim. 31
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The suggestions proved to be interesting to the programmers and helpful in resolving the code errors.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.3. Error Checking and Debugging in Code
plaindirect- 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.
It may state two separate claims as one.
- Rests on an assumption: The authors' assessment that ChatGPT's debugging suggestions were helpful reflects actual programmer benefit.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Of the participating students, 61.3% (35.5% first year, 16.1% second year, 6.5% third year, and 3.2% fourth year) were undergraduate students, 32.2% (16.1% first year and 16.1% second year) were Master's students, and 6.5% were second-year doctoral students.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 5. Survey on ChatGPT Language Model for Programming Learning and Teaching Support
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“ChatGPT can identify errors in code and provide potential suggestions and code snippets.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.3. Error Checking and Debugging in Code
plaindirect- 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.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, when we asked ChatGPT to “provide a pseudocode for a selection sort and an explanation”, it generated the pseudocode and explanation, as depicted in Figure 7.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.1. Conceptual Understanding
plaindirectSingle case- 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.
Probably not a substantial claim.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The generated optimized code was validated on the AOJ platform, and it passed all test cases (10/10) and was accepted for the Bubble Sort problem”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.4. Solution Code Optimization
plainSingle case- 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.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“ChatGPT can also be used to generate topic-specific quiz questions tailored to the subject matter and difficulty level.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Lesson Planning:
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“ChatGPT can generate nearly accurate answers to technical questions from a wide range of topics and correct or partially correct programming code based on problem descriptions, algorithm and problem names, etc.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
plaindirect- 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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“the experimental results show that the accuracy of the generated code is 95.83% on the basic compiler and 75% on AOJ.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 7. Limitations
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.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“ChatGPT can be used to create lesson plans for specific courses, such as math, chemistry, physics, computer science, civil engineering, language, and literature.ChatGPT provides topic-specific illustrations, activities, and exercises to help educators better teach their students.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Lesson Planning:
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our surveys and experimental results show that ChatGPT is useful not only for programming education but also for education and research.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 8. Conclusions
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.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“During the experiments, we obtained significant results in code generation, error checking and debugging, and optimization of the solution code.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 7. Limitations
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.
It may state two separate claims as one.
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
“This is a rare case (or an example of erroneous code), but ChatGPT has optimized and fixed it.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4.4. Solution Code Optimization
plaindirectSingle 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.
It may state two separate claims as one.
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
“This can lead to social unrest, reputation damage, and even physical harm.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
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
“ChatGPT is a revolutionary LLM that can maintain humanlike conversations and generate human-like text for any natural language query that is nearly indistinguishable.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 8. Conclusions
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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
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
“Simply acquiring answers and code from ChatGPT can be a barrier to improving learners' critical thinking and problem-solving skills.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
plaindirect- 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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“However, the quality and suitability of these applications for programming learning remain unclear.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 4. Programming Learning with ChatGPT
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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no 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
“the solution codes generated by AI models can be used for academic coding exams and competitions”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no 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
“Educators can use ChatGPT to provide personalized learning support for their students.Depending on a student's needs and learning style, ChatGPT can suggest customized resources and learning activities.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Lesson Planning:
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no 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
“This is because the ease of obtaining answers, problem-solving strategies, and scientific text generation can limit critical thinking and problem-solving skills.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Blind reliance on generative AI tools:
plaindirect- 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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no 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
“The model can be used to answer questions, write essays, solve problems, explain complex topics, provide virtual tutoring, practice languages, learn programming, teach, and support research.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 8. Conclusions
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.
It may state two separate claims as one.
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
“the ethical implications (e.g., bias and discrimination, privacy and security, misuse of technology, accountability, transparency, and social impact) of ChatGPT are complex and multifaceted”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 8. Conclusions
plain- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no 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
“it can be difficult to hold anyone accountable for the responses generated”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
plain- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
- Rests on an assumption: ChatGPT generates responses without human intervention.
In the paper
“Cotton et al. [57] presented several strategies for recognizing texts from LLMs such as ChatGPT, including language inconsistencies, a lack of proper citations, factual errors, ambiguity, and poor context awareness.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > -Sam Altman, CEO of OpenAI
What the sources say
- Not checked yetcites [57] Cotton, D.R., Cotton, P.A., and Shipway, J.R. (2023). Chatting and Cheating. Ensuring academic integrity in the era of ChatGPT. Innov. Educ. Teach. Int., 1–12. source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
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
“ChatGPT and other generative models involve the collection and processing of personal data, which raises concerns about privacy and data security.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
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
“If the answer is identical to answers generated by ChatGPT, it may be a sign that the answer was not generated by a human.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Compare responses:
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“The answers and texts generated by ChatGPT are becoming indistinguishable from human-generated answers and texts.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > -Sam Altman, CEO of OpenAI
What the sources say
- Not checked yetcites [57] Cotton, D.R., Cotton, P.A., and Shipway, J.R. (2023). Chatting and Cheating. Ensuring academic integrity in the era of ChatGPT. Innov. Educ. Teach. Int., 1–12. source
- Not checked yetcites [60] Elkins, 2020, Can GPT-3 pass a Writer’s turing test?, J. Cult. Anal.
- Not checked yetcites [61] Gao, C.A., Howard, F.M., Markov, N.S., Dyer, E.C., Ramesh, S., Luo, Y., and Pearson, A.T. (2022). Comparing scientific abstracts generated by ChatGPT to original abstracts using an artificial intelligence output detector, plagiarism detector, and blinded human reviewers. bioRxiv. source
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.
Probably not a substantial claim.
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
“In short, ChatGPT threatens the fairness and validity of online exams and assignments.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 6. Threats and Strategies
plaindirect- 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.
Probably not a substantial claim.
- Rests on an assumption: ChatGPT can generate human-like text for academic topics.
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
“Furthermore, the ChatGPT model can be used to solve technical (e.g., engineering and computer programming) and non-technical (e.g., language and literature) problems.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 8. Conclusions
plaindirect- 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.
It may state two separate claims as one.
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
“First, it can effectively support the writing process of research.At its most basic, it can improve writing by finding and correcting typographical errors, improving grammatical inconsistencies, providing advanced vocabulary, and recommending improvement strategies.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 3.3. Opportunities for Researchers
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
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
“ChatGPT and other AI language models can be used to generate fake news, hate speech, and other harmful content.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
plaindirect- 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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
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
“This can help detect cases of fraud.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Compare responses:
plainconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
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
“ChatGPT responses typically have certain characteristics, such as a lack of personalization or a rather generic tone.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no 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
“Educators can receive the help of ChatGPT to answer learners' questions.Furthermore, if educators can ask ChatGPT for explanations and examples on a particular topic, it will surely increase the effectiveness of teaching.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Lesson Planning:
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
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
“If the answers appear disjointed or nonsensical, this may indicate that they were generated by ChatGPT or another AI model.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no 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
“ChatGPT responses may not have a consistent or logical flow, especially when it is generating answers to complex questions.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no 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
“The model can also summarize published work on a particular topic, which helps researchers understand the work.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 3.3. Opportunities for Researchers
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“To the best of our knowledge, there are no such tools that can recognize code generated by AI models”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Ethical implications and potential biases:
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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no 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
“it still has certain limitations, such as a lack of common sense, potential bias, difficulty with complex reasoning, and inability to process visual information.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 8. Conclusions
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: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
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
“The model can be used to check submitted assignments for plagiarism.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > Rapid Assessment and Evaluation:
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“It is worth noting that our experiments were conducted on ChatGPT, which is currently in an active development phase.”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 7. Limitations
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
The paper relies on this without support.
The eight randomly selected AOJ problems and three code generations per problem are representative of ChatGPT's general code-generation performance for programming learning.
“We leveraged the problem descriptions of eight random problems and generated codes three times for each problem on the basis of the same description.”
The paper relies on this without support.
The survey respondents are representative of students and teachers who use ChatGPT for programming learning and teaching.
“We conducted a comprehensive survey on how ChatGPT supports programming learning and teaching among undergraduate, Master's, and doctoral students and teachers.”
The paper relies on this without support.
Students' self-reported helpfulness of ChatGPT suggestions reflects actual improvement in solving programming problems.
“By asking the question “Do ChatGPT's suggestions help you in solving programming problems?”,we tried to assess how ChatGPT suggestions help students.We found that about 86.7% of the students answered “Yes”.”
The paper relies on this without support.
The authors' judgment that ChatGPT-generated explanations and pseudocode are easy to understand and useful reflects learners' actual understanding.
“The results demonstrate that ChatGPT can generate easy-to-understand explanations and pseudocode that are useful for learners to understand algorithmic concepts.”
The paper relies on this without support.
The authors' assessment that ChatGPT's debugging suggestions were helpful reflects actual programmer benefit.
“The suggestions proved to be interesting to the programmers and helpful in resolving the code errors.”
The paper relies on this without support.
ChatGPT can generate human-like text for academic topics.
“As ChatGPT can generate human-like text for academic topics”
The paper relies on this without support.
LLMs rely heavily on training data, and when the data contain biases or anomalies, it could lead to unfair results.
“Typically, LLMs rely heavily on training data, and when the data contain biases or anomalies, it could lead to unfair results.”
The paper relies on this without support.
The internal mechanisms and processes of ChatGPT are not sufficiently open and transparent to users about how they work.
“Furthermore, the internal mechanisms and processes are not sufficiently open and transparent to users about how they work.”
The paper relies on this without support.
ChatGPT generates responses without human intervention.
“Because ChatGPT generates responses without human intervention”
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 check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“Particularly, students that are more experienced using ChatGPT tend to be less concerned about its impact on academic integrity and more optimistic about its role in self-study (Nisar and Aslam 2023;Rahman and Watanobe 2023).”
From Widely used but barely trusted: understanding student perceptions on the use of generative AI in higher education (2025) · cites this paper as Rahman and Watanobe 2023
Experimental — accuracy not yet measuredIn the citing paper
“foster students’ performance, engagement and motivation [14, 44, 58], and enhance teaching practices [17, 18, 64, 74].”
From A systematic literature review of empirical research on ChatGPT in education (2024) · cites this paper as [58]
In the citing paper
“fostering students' performance, engagement, and motivation, leading to increase students' academic success [14, 44, 58]”
From A systematic literature review of empirical research on ChatGPT in education (2024) · cites this paper as [58]
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
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > 8. Conclusions
“Simon et al. [25]”
§ ChatGPT for Education and Research: Opportunities, Threats, and Strategies > References
“25. Frieder, S.; Pinchetti, L.; Griffiths, R.R.; Salvatori, T.; Lukasiewicz, T.; Petersen, P.C.; Chevalier, A.; Berner, J. Mathematical capabilities of chatgpt.”
Citations
Cites
1 retracted
Retracted: Chat With ChatGPT on Intelligent Vehicles: An IEEE TIV Perspective
Du, H., Teng, S., Chen, H., Ma, J., Wang, X., Gou, C., Li, B., Ma, S., Miao, Q., Na, X., et al. (2023) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2023.3253281 · Cited by 164 · cited 1×
“Moreover, ChatGPT has recently been applied to tasks in different domains such as software engineering [36], medical education [37], data augmentation [38], fifth industrial revolution [39], code generation [40], intelligent vehicles [41], and solving AI tasks [42].”
MethodRetractedVaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 1×
“Therefore, the Transformer architecture was introduced [4] to alleviate this problem.”
MethodLanguage Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372 · cited 1×
“Generative pretrained transformer (GPT)-3 is a large language model (LLM) based on the Transformer architecture [5,6] that has achieved significant success in NLP tasks.”
BackgroundChatGPT for Good? On Opportunities and Challenges of Large Language Models for Education
Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · doi:10.35542/osf.io/5er8f · Cited by 7,552
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326 · cited 3×
“Furthermore, several studies [9–12 Frieder et al. [25] tested the mathematical capabilities of ChatGPT on GHOSTS and handcrafted datasets.”
BackgroundMethodTraining language models to follow instructions with human feedback
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C. L., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. (2022) · arXiv · doi:10.48550/arxiv.2203.02155 · Cited by 4,208 · cited 2×
“Finally, the resulting model is called InstructGPT [8].”
BackgroundMethodKung, T. H., Cheatham, M., ChatGPT, Medenilla, A., Sillos, C., De Leon, L., Elepaño, C., Madriaga, M., Aggabao, R., Diaz-Candido, G., et al. (2022) · doi:10.1101/2022.12.19.22283643 · Cited by 4,115
Kung, T. H., Cheatham, M., Medenilla, A., Sillos, C., De Leon, L., Elepaño, C., Madriaga, M., Aggabao, R., Diaz-Candido, G., Maningo, J., et al. (2023) · PLOS Digital Health · doi:10.1371/journal.pdig.0000198 · Cited by 4,115 · cited 1×
“Kung et al. [26] evaluated the performance of ChatGPT on the US Medical Licensing Exam, which includes three exams: Step 1, Step 2CK, and Step 3.”
BackgroundChatting 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 3×
“Students can be given clear instructions on assignments and online exams on how to structure their assignments and answer their questions online [57].”
BackgroundChatting and Cheating. Ensuring academic integrity in the era of ChatGPT
Cotton, D., Cotton, P., & Shipway, J. R. (2023) · doi:10.35542/osf.io/mrz8h · Cited by 3,079
Research works printed but not linked: 3
- Rahman, M.M. (2022). Data Analysis and Code Assessment Using Machine Learning Techniques for Programming Activities. [Ph.D. Thesis, The University of Aizu].
- Avila-Chauvet, L., Mejía, D., and Acosta Quiroz, C.O. (2023). Chatgpt as a Support Tool for Online Behavioral Task Programming. SSRN, Available online: https://ssrn.com/abstract=4329020.
- Jiao, W., Wang, W., Huang, J.t., Wang, X., and Tu, Z. (2023). Is ChatGPT a good translator? A preliminary study. arXiv.
Cited by
Sorted by citations
Items 501–600 of 1,004
Curse or a Blessing: Excessive Use of ChatGPT in Academia
Sajjad, M. (2024) · Bulletin of Business and Economics (BBE) · doi:10.61506/01.00329 · Cited by 7
“Whilc ChatGPT offered very convenient and efficient way to offer the knowledge to the students at higher level, various concerns have been raised for its excessive use which started a debate that either the excessive use of ChatGPT in academia is a blessing or a curse (Rahman & Watanobe, 2023).”
BackgroundThe Use of ChatGPT in Improving and Reviewing Scientific Paper Writing: An Exploratory Study
Bernardino, M., Cargnelutti, R., Garcia, R. D. S., & Silva, W. (2024) · IEEE Revista Iberoamericana de Tecnologias del Aprendizaje · doi:10.1109/rita.2024.3458848 · Cited by 6
“On the other hand, researchers expressed concerns regarding the ethical issues related to ChatGPT as well as its potential adverse effects, including the possibility of cheating on online exams, the impact on critical thinking skills, and the difficulties in evaluating the generated information [8].”
BackgroundReal Customization or Just Marketing: Are Customized Versions of Generative AI Useful?
Garrido‐Merchán, E. C., Barrigüete, J. L. A., Borrás-Pala, F., Escobar-Torres, L., Zorita, C. M. D. I., Ortiz-Lozano, J. M., & Vieites, A. R. (2024) · F1000Research · doi:10.12688/f1000research.153129.3 · Cited by 5
The Role of ChatGPT and Artificial Intelligence in Education
Martínez, M. Á. Q., Tumaille-Quintana, D.-S., Moran-Burgos, A.-D., & Ríos, M. D. G. (2024) · IEEE Colombian Conference on Communications and Computing (COLCOM) · doi:10.1109/colcom62950.2024.10720308 · Cited by 5
“Artificial intelligence has the potential to completely change education by providing more personalized and effective learning [18], [11], [19], [30], [3], [27], [10], [15], [2], [4], [17], [23].”
BackgroundDeveloping written communication skills in engineers in Spanish: is ChatGPT a tool or a hindrance?
Mirón-Mérida, V. A., & García‐García, R. (2024) · Frontiers in Education · doi:10.3389/feduc.2024.1416152 · Cited by 5
Hong, S.-J., Lee, Y., & Kim, S.-W. (2024) · International Journal on Advanced Science Engineering and Information Technology · doi:10.18517/ijaseit.14.6.10092 · Cited by 10
Šumak, B., Pušnik, M., Kožuh, I., Šorgo, A., & Brdnik, S. (2025) · Applied Sciences · doi:10.3390/app15020631 · Cited by 6
Enhancing Mathematics Education with ChatGPT-4 Personalized Problem-Solving and Consistent Learning
Chohan, M. I., & Khan, I. M. (2024) · International Conference on Computing and Data Analytics (ICCDA) · doi:10.1109/iccda64887.2024.10867401 · Cited by 5
From the Gender Lens: Student Perceptions of ChatGPT in Higher Education
Fadillah, M. A., & Akbar, M. F. (2025) · Advances in Mobile Learning Educational Research · doi:10.25082/amler.2025.01.015 · Cited by 8
ChatGPT in Research and Education: A SWOT Analysis of Its Academic Impact
Miah, A. S. M., Tusher, M. M. R., Hossain, M. M., Hossain, M. M., Rahim, M. A., Hamid, M. E., Islam, M. S., & Shin, J. (2025) · Computer Modeling in Engineering & Sciences · doi:10.32604/cmes.2025.064168 · Cited by 11
“Rahman et al. [34] analyze how ChatGPT aids learners in developing programming and problem-solving skills.”
InfluentialDoes the Improvement in AI Tools Necessitate a Different Approach to Engineering Education?
Hauge, J. B., & Jeong, Y. (2024) · Advances in transdisciplinary engineering · doi:10.3233/atde240211 · Cited by 6
Implementing AI in Physics Lessons in the High School
Bessas, N., Tzanaki, E., Vavougios, D., & Plagianakos, V. P. (2023) · International Conference on Computational Science and Computational Intelligence (CSCI) · doi:10.1109/csci62032.2023.00293 · Cited by 5
“In fact, it has an advantage over similar LLM models as it displays a more realistic character [12].”
BackgroundLearning with Conversational AI: ChatGPT and Bard/Gemini in Education
da Silva, G. S., & Ulbricht, V. R. (2024) · Cognition and exploratory learning in the digital age · doi:10.1007/978-3-031-66462-5_6 · Cited by 6
Critical Thinking in the Age of Algorithms
Andrae, S. (2024) · Advances in educational marketing, administration, and leadership book series · doi:10.4018/979-8-3373-1692-5.ch010 · Cited by 5
Dobre, S.-C., & Popescu, E. (2024) · International Conference on Information Technology Based Higher Education and Training (ITHET) · doi:10.1109/ithet61869.2024.10837650 · Cited by 7
“Paper [14] addresses the opportunities, challenges, and strategies of using ChatGPT in education.”
Baskara, F. R. (2024) · Proceeding of The International Conference of Inovation Science Technology Education Children and Health · doi:10.62951/icistech.v4i1.80 · Cited by 5
A Study on Students’ Critical Thinking Cultivation in English Teaching in the ChatGPT Era
Liu, Y., & Wei, L. (2024) · Advances in Social Science, Education and Humanities Research/Advances in social science, education and humanities research · doi:10.2991/978-2-38476-263-7_10 · Cited by 5
ChatGPT to enhance learning in dental education at a historically black medical collegePreprint
Rahad, K., Martin, K., Amugo, I., Ferguson, S., Curtis, A., Davis, A., Gangula, P. R., & Wang, Q. (2023) · Research Square · doi:10.21203/rs.3.rs-3546693/v2 · Cited by 5
“Comparing with other educational resources, ChatGPT demonstrates clear advantages in terms of cost, speed, accuracy, customization, user-friendliness, seamless integration with existing tools, and scalability [3].”
BackgroundNyaaba, M., Min, S., Apam, M. A., Acheampong, K. O., Dwamena, E., & Zhaı, X. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5174910 · Cited by 10
Artificial Intelligence in Enhancing Korean High School Students’ English Skills
Kim, R. (2025) · New language learning and teaching environments · doi:10.1007/978-3-031-83561-2_5 · Cited by 5
Generative AI Integrated Educational Model for User-Centered Design
Wu, Y., Zeng, X., & Lin, Q. (2024) · doi:10.1145/3722237.3722260 · Cited by 4
Exploring Teachers’ Beliefs About ChatGPT in Arts Education
Kladaki, M., Κώστας, Α., & Alexopoulos, P. (2025) · Education Sciences · doi:10.3390/educsci15070795 · Cited by 7
How Generative AI Impact Student Emotion and Engagement in Programming Tasks?
Ma, B., Guo, L., Yang, T., & Ding, J. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-98462-4_30 · Cited by 4
The impact of LLMs on mathematics education and research at the university
Rizos, I., & Gkrekas, N. (2025) · Social Sciences & Humanities Open · doi:10.1016/j.ssaho.2025.101969 · Cited by 8
Tran, N.-M., & Le, V.-T. (2025) · Technical Education Science/Giáo dục Kỹ thuật · doi:10.54644/jte.2025.1738 · Cited by 7
Large language models as a catalyst for opportunity recognition capability
Renz, P. A., Martin, J., & Flad, M. (2025) · International Journal of Entrepreneurial Behaviour & Research · doi:10.1108/ijebr-08-2024-0815 · Cited by 7
Cho, N., Li, H., Zhang, X., & Li, Y. (2025) · Journal of Science Education and Technology · doi:10.1007/s10956-025-10280-7 · Cited by 5
“Among the myriad of technological advances, computer simulations and generative artificial intelligence (GenAI) stand out as significant innovations with the potential to revolutionize science education (Rahman & Watanobe, 2023; Cooper, 2023).”
Lin, P., Deng, Q., & Zhou, Y. (2026) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2025.100534 · Cited by 4
Büyükeke, A. (2025) · Üniversite Araştırmaları Dergisi · doi:10.32329/uad.1728785 · Cited by 4
Sığırtmaç, İ. C., Türkmen, C., Değerli, Y. İ., & Öksüz, Ç. (2026) · Interactive Learning Environments · doi:10.1080/10494820.2026.2617984 · Cited by 5
Wani, T. A., Liem, M. K., Prasad, N., Robinson, K. M., Nexhip, A., Tassos, M., Gjorgioski, S., Khan, U. R., Boyd, J., & Riley, M. M. (2026) · JMIR Medical Education · doi:10.2196/82988 · Cited by 4
“These include ChatGPT’s use of inaccurate content (including fictitious reference material) [25], concerns regarding human teacher replacement [11], and negative repercussions on students’critical thinking and problem-solving skills [26].”
Gambo, O. O., Onifade, O. J., Olanrewaju-Smart, W., Abuya, O. J., Salawu, R. E., Massenon, R., & Gambo, I. (2026) · Discover Education · doi:10.1007/s44217-026-01142-4 · Cited by 10
“For instance, Rahman and Watanobe [77] found that students who used ChatGPT demonstrated improved debugging skills, however, the tool's overall problem-solving capabilities remained limited, particularly in more complex tasks.”
Jang, H., Jin, L., & So, H.-J. (2026) · The Asia-Pacific Education Researcher · doi:10.1007/s40299-025-01076-9 · Cited by 4
“Learners expressed frustration with ChatGPT’s lack of evidence-based support and unclear sources, echo - ing concerns raised in recent literature (Kasneci et al., 2023 ; Rahman & Watanobe, 2023; Wang, 2024 ).”
Chen, M.-Y., & Hung, S.-Y. (2026) · European Journal of Educational Research · doi:10.12973/eu-jer.15.2.579 · Cited by 8
Xu, J., Sun, D., Feng, Y., Shadiev, R., & Li, Y. (2026) · International Journal of STEM Education · doi:10.1186/s40594-026-00606-1 · Cited by 6
Wang, D., & Wang, Z. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1794730 · Cited by 4
Advantages and Disadvantages of Chatgpt in Science Learning: A Systematic Literature Review
Asmiddin, A. M., Nurhuda, P., & Megawati, R. (2023) · Jurnal Penelitian Pendidikan IPA · doi:10.29303/jppipa.v9i12.6576 · Cited by 6
“…(Farrokhnia et al., 2023b) Speed and are very responsive in science learning (Xiao & Zhi, 2023b; Yilmaz & Karaoglan Yilmaz, 2023); (Zheng, 2023); (Rahman & Watanobe, 2023); (Bitzenbauer, 2023); (Cooper, 2023b); (Anagnostopoulos, 2023); (Mondal et al., 2023); (Ivanov & Soliman, 2023); (AlZaabi et…”
ChatGPT and Higher Education: A Pathway to Unprecedented Progress
Almahasees, Z. M., Awabdeh, A. A., & Mahmoud, S. (2024) · Journal of Higher Education Theory and Practice · doi:10.33423/jhetp.v24i5.7004 · Cited by 10
“Rahman and Watanobe (2023) highlight the role, opportunities, threats, and strengths of CHATGPT for education.”
BackgroundOsunbunmi, I. S., Cutler, S. L., Dansu, V., Brijmohan, Y., Bamidele, B., Udosen, A. N., Arinze, L., Oje, A. V., Moyaki, D., Hicks, M., et al. (2024) · ASEE Annual Conference & Exposition · doi:10.18260/1-2--47041 · Cited by 6
Educational Innovation in the Information Age
Mrabet, J., Studholme, R., & Thompson, N. L. (2024) · Advances in human resources management and organizational development book series · doi:10.4018/979-8-3693-3096-8.ch013 · Cited by 4
Galamiton, N. S., Bacus, S., Fuentes, N., Ugang, J., Villarosa, R., Wenceslao, C., & Ocampo, L. A. (2024) · International Journal of Computational Intelligence Systems · doi:10.1007/s44196-024-00668-5 · Cited by 6
“In the IT sector, LLMs assistance was also tested in improving programming skills [46].”
Interactive Online Education Using LLMs: Foundations and Applications
Xie, Z., Huang, J., Xing, R., & Li, Q. (2024) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-981-97-9255-9_1 · Cited by 4
Hooper, K., & Lunn, S. J. (2024) · IEEE Frontiers in Education Conference (FIE) · doi:10.1109/fie61694.2024.10893167 · Cited by 4
“AI can impact academia in areas ranging from help with academic writing for students to potential opportunities for evaluation and assessment for instructors [21], [22].”
Oh, S.-Y., Cao, Y. L., Katz, A., & Zhao, J. (2024) · IEEE Frontiers in Education Conference (FIE) · doi:10.1109/fie61694.2024.10893102 · Cited by 7
“The ability of AI to generate tailored educational content and engage students in deeper learning processes presents a promising avenue to enhance educational outcomes [4, 5].”
InfluentialHeriyawati, D. F., & Romadhon, M. G. E. (2025) · Computer-Assisted Language Learning Electronic Journal · doi:10.54855/callej.252614 · Cited by 6
AliceBot: Designing AI-Driven Applications for Mental Health Management
Zhao, M., Bai, Y., Zhong, W., Cun, W., Li, Y., Zhu, M., Feng, Y., & Chen, D. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-93712-5_16 · Cited by 4
Shanto, S. S., Ahmed, Z., & Jony, A. I. (2024) · International Conference on Computing Advancements (ICCA) · doi:10.1145/3723178.3723268 · Cited by 4
Rahman, A., Islam, M. K., Al-Mamun, A., & Islam, M. S. (2025) · F1000Research · doi:10.12688/f1000research.161351.1 · Cited by 8
Chen, B., & Chen, C. (2023) · IEEE International Conference on Knowledge Innovation and Invention (ICKII) · doi:10.1109/ickii58656.2023.10332623 · Cited by 4
DİNÇER, R., Dinçer, N., & GUKSU, O. (2023) · The Literacy Trek · doi:10.47216/literacytrek.1382543 · Cited by 5
ChatGPT and Generative AI in Educational Ecosystems
Haidar, A. (2024) · Advances in educational marketing, administration, and leadership book series · doi:10.4018/979-8-3693-1536-1.ch004 · Cited by 4
Future Office: A Comparative Study on the Acceptance and Utilization of Generative AI Technologies
Bies, L. M., Schmidt, S., Morana, S., & Werth, D. (2024) · International Conference on Artificial Intelligence, Computer, Data Sciences and Applications (ACDSA) · doi:10.1109/acdsa59508.2024.10467651 · Cited by 6
“Capable of engaging in human-like conversations and tackling complex challenges, ChatGPT raises the central question of whether Generative AI can revolutionize business operations, offering avenues for enhanced efficiency and innovation [2].”
BackgroundReal Customization or Just Marketing: Are Customized Versions of Generative AI Useful?
Garrido‐Merchán, E. C., Barrigüete, J. L. A., Borrás-Pala, F., Escobar-Torres, L., Zorita, C. M. D. I., Ortiz-Lozano, J. M., & Vieites, A. R. (2024) · F1000Research · doi:10.12688/f1000research.153129.1 · Cited by 11
ETHICAL CONCERNS IN USING OF GENERATIVE TOOLS IN HIGHER EDUCATION: CROSS - COUNTRY STUDY
Mironova, J., Riashchenko, V., Kinderis, R., Djakona, V., & Dimitrova, S. (2024) · Environment Technology Resources Proceedings of the International Scientific and Practical Conference · doi:10.17770/etr2024vol2.8097 · Cited by 4
“The impact of AI extends far and wide, triggering discussions on diverse topics, from its potential to transform learning and teaching methods [13] to its role in research and the imperative considerations of ethics and academic integrity [8].”
BackgroundApplications, Challenges and Early Assessment of AI and ChatGPT in Education
Sidiropoulos, D., & Anagnostopoulos, C. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-65996-6_1 · Cited by 40
Investigating Advanced Generative Dialogue Systems for Educational Chatbots
Binhammad, M. H. Y., Othman, A., Abuljadayel, L., Mheiri, H. A., Alkaabi, M., & Almarri, M. (2024) · Creative Education · doi:10.4236/ce.2024.158096 · Cited by 7
Evaluating the Use of ChatGPT by University and High School Students
Szügyi, J. S., Szügyi, E., & Takács, M. (2024) · IEEE International Symposium on Applied Computational Intelligence and Informatics (SACI) · doi:10.1109/saci60582.2024.10619721 · Cited by 5
“Rahman [12] asks university students and teachers about their use of ChatGPT in learning programming.”
BackgroundCultivating Software Quality Improvement in the Classroom: An Experience with ChatGPT
AlOmar, E. A., & Mkaouer, M. W. (2024) · IEEE International Conference on Software Engineering Education and Training (CSEE&T) · doi:10.1109/cseet62301.2024.10663028 · Cited by 6
“Therefore, the use of ChatGPT in education has become an area of debate about its opportunities and threats to student learning [18].”
BackgroundSánchez-Reina, J. R., Theophilou, E., Hernández‐Leo, D., & Ognibene, D. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-67351-1_26 · Cited by 5
“The use of AI tools such as ChatGPT in educational settings can help support students in diverse learning materials, such as coding [22] and microbiology [2].”
MethodFichtner, U. A., Knaus, J., Graf, E., Koch, G., Sahlmann, J., Stelzer, D., Wolkewitz, M., Binder, H., & Weber, S. (2024) · PLoS ONE · doi:10.1371/journal.pone.0308375 · Cited by 4
Navigating Educational Leadership Research With Generative Artificial Intelligence
LaFrance, J. A. (2024) · Advances in educational marketing, administration, and leadership book series · doi:10.4018/979-8-3693-6220-4.ch006 · Cited by 4
Rosas-Meléndez, S. A., Chans, G. M., López-Velázquez, P. M., Álvarez-Siordia, F. M., & Camacho-Zúñiga, C. (2024) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-981-97-9255-9_13 · Cited by 4
Adhantoro, M. S., Gunawan, D., Prayitno, H. J., Riyanti, R. F., Purnomo, E., & Jufriansah, A. (2025) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2025.1446590 · Cited by 9
AI Generated Code Plagiarism Detection in Computer Science Courses: A Literature Mapping
Simmons, A., de Holanda, M. T., Chamon, C., & Silva, D. D. (2024) · IEEE Frontiers in Education Conference (FIE) · doi:10.1109/fie61694.2024.10893117 · Cited by 6
Braga, V. (2023) · Journal of Engineering Science · doi:10.52326/jes.utm.2023.30(3).07 · Cited by 3
“The capacity of LLM’s to support self-learning for students is widely known already [26–28] , yet the academia is perceiving at the moment these opportunities more like a threat to the classical educational approach [29].”
BackgroundDeviant acts in the use of ChatGPT: An analytical study of student behaviour
Purwasih, W., & Sahnan, A. (2023) · INSANIA Jurnal Pemikiran Alternatif Kependidikan · doi:10.24090/insania.v28i2.9529 · Cited by 7
“…(Setiawan & Luthfiyani, 2023; Supriyadi, 2022; Waluyo et al., 2023), challenges of using ChatGPT in research and education (Faiz & Kurniawaty, 2023; Rahman & Watanobe, 2023), educational institutions need to create new policies to follow up on advances in ChatGPT technology to maintain academic…”
BackgroundAdvantages, Challenges and Prospects of ChatGPT in Oral English Teaching
Shi, X. (2024) · Transactions on Social Science Education and Humanities Research · doi:10.62051/c49b2t84 · Cited by 6
“Rahman and Watanobe [20] also discuss opportunities, threats, and strategies for ChatGPT in education and research.”
BackgroundUnveiling Chat GPT’s Educational Prospects: A SWOT Analysis
Dar, M. A., Khursheed, T., Ahmad, A., & Fayaz, R. (2024) · International Conference on Computing for Sustainable Global Development · doi:10.23919/indiacom61295.2024.10499038 · Cited by 6
Abhishek, N., Devesh, S., Dinesh, S., Kulal, A., & Suraj, N. (2023) · Research Square · doi:10.21203/rs.3.rs-3321483/v1 · Cited by 5
Integrating AI into Design Ideation: Assessing ChatGPT’s Role in Human-Centered Design EducationPreprint
Papachristos, E., İnal, Y., Vicient, C., Johansen, E. A., & Hermansen, M. (2024) · doi:10.36227/techrxiv.171656320.09963657/v1 · Cited by 5
A Theoretical Framework of Student Agency in AI-Assisted Learning
Dai, Y., & Lai, S. (2025) · Oxford University Press eBooks · doi:10.1093/9780198945253.003.0009 · Cited by 6
“Since the launch of tools such as ChatGPT, higher education students have increasingly adopted GenAI tools to assist their learning tasks, including writing, coding, data analysis, and receiving feedback (Lai et al., 2025; Rahman & Watanobe, 2023).”
Cabrera, C., & Neville, R. (2025) · Perspectives: Policy and Practice in Higher Education · doi:10.1080/13603108.2025.2595453 · Cited by 4
Generative Artificial Intelligence Enables Multiple Learning Environments and Implementation Paths
Jiang, R., He, X., Zhang, S., Zhou, Y., & Han, Y. (2024) · International Conference on Educational Technology Management (ICETM) · doi:10.1145/3711403.3711436 · Cited by 3
“In computer science, AI tools have shown great potential in coding education, as evidenced by GPT-3’s ability to generate different code interpretations [12].”
Research trends and the impact of ChatGPT on educational environments
Firdaus, T., Mufidah, R., Hamida, R. N., Febrianti, R. I., & Guivara, A. E. R. (2025) · International Journal of Social Sciences and Education Research · doi:10.24289/ijsser.1609741 · Cited by 6
ChatGPT na Universidade de Perugia: Potencializando a Interação entre Docentes e Discentes
Brognoli, P. C., & Rodrigues, M. B. (2025) · Encontros Bibli Revista Eletrônica de Biblioteconomia e Ciência da Informação · doi:10.5007/1518-2924.2025.e102135 · Cited by 3
Tran, H. V., Phuong, L. V. T., Cong, Q. N., & Tam, N. T. L. (2025) · Interactive Learning Environments · doi:10.1080/10494820.2025.2511238 · Cited by 3
“A favorable educational environment, where teachers and schools encourage the application of new technologies such as ChatGPT, will strongly motivate students to make the most of this tool (Rahman & Watanobe, 2023).”
Bernstein, E., Ramsamooj, A., Millar, K. L., & Lum, Z. C. (2025) · JMIR AI · doi:10.2196/68603 · Cited by 3
“However, the potential for academic dishonesty and the ethical implications of AI use in education remain significant concerns [48,53,54,59,67,71].”
Lahmer, M. (2025) · Journal of Digital Pedagogy · doi:10.61071/jdp.2531 · Cited by 3
Intention to use ChatGPT among law educators in Saudi Arabia
Sarabdeen, J. (2025) · Frontiers in Education · doi:10.3389/feduc.2025.1631413 · Cited by 4
Perera, P., Kulathunga, V. K., Perera, N., Gunarathna, A., Jayasinghe, P., & Rajamanthri, L. (2025) · Asian Education and Development Studies · doi:10.1108/aeds-11-2024-0275 · Cited by 7
Bancalero, P. D. L. F., Guillén‐Gámez, F. D., Stošić, L. V., & Giménez-Gualdo, A. M. (2025) · IGI Global eBooks · doi:10.4018/979-8-3373-7912-8.ch005 · Cited by 3
Portuguese secondary school students’ perceptions regarding the use of ChatGPT
Callado, A., Saúde, I., & Araújo, J. L. (2025) · Contemporary Educational Technology · doi:10.30935/cedtech/17516 · Cited by 9
Soyoof, A., Reynolds, B. L., Rassaei, E., Kao, C.-W., & Ha, X. V. (2025) · Computers and Education Artificial Intelligence · doi:10.1016/j.caeai.2025.100530 · Cited by 7
Wei, L., Qiu, L., & Man, S. (2025) · SAGE Open · doi:10.1177/21582440251407506 · Cited by 5
“AI becomes an appealing option by providing quick solutions to academic queries, improving short-term performance (Rahman & Watanobe, 2023).”
Can ChatGPT as a generative AI tool enhance university student academic achievement?
Yu, Q., Yu, K., Yang, R., & Li, B. (2025) · Innovations in Education and Teaching International · doi:10.1080/14703297.2025.2605651 · Cited by 5
Towards Accessible Information Retrieval for Children With a Mild Intellectual Disability
Weijers, R., Ooms, S., Pelrine, K., & Hauptmann, H. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-12717-4_9 · Cited by 3
Salman, H., Ahmad, M. A., Ibrahim, R. B., & Mahmood, J. (2026) · PeerJ Computer Science · doi:10.7717/peerj-cs.3417 · Cited by 3
Liu, C., & Liang, M. (2026) · System · doi:10.1016/j.system.2026.104018 · Cited by 7
Hybrid Intelligence Feedback: A Model Proposal and Preliminary Study
Coşkun, A., & Er, E. (2026) · Journal of Computer Assisted Learning · doi:10.1002/jcal.70212 · Cited by 3
Kibria, M. G., Khanom, K., Sultana, J., & Chowdhury, F. P. (2026) · Global Business Review · doi:10.1177/09721509261443740 · Cited by 5
Su, F., Beh, W. F., & Lee, K. C. S. (2026) · Acta Psychologica · doi:10.1016/j.actpsy.2026.107353 · Cited by 7
ChatGPT's Technology Application in the Higher Education Sector, Risk Analysis and Pathway Changes
Ke, Y. (2023) · Journal of Education and Educational Research · doi:10.54097/jeer.v3i3.9635 · Cited by 3
“With the help of technology, teachers and students are expected to cross the information divide previously formed by geographical differences and lack of technology, rapidly improve teachers' education and teaching, and enhance students' information literacy [2].”
Background8th Virtual International Conference on Education, Innovation and ICT
Varii, A. (2024) · Adaya Press eBooks · doi:10.58909/adc24139168 · Cited by 3
The aid of ChatGPT to dance education: a theoretical exploration based on TPACK
Dou, X. (2024) · Research in Dance Education · doi:10.1080/14647893.2024.2371897 · Cited by 3
Gezgin, D. M., Mert, S., Kesici, A. İ., & Yıldırım, S. (2024) · Sakarya University Journal of Education · doi:10.19126/suje.1426980 · Cited by 4
The Effect of ChatGPT on Education in the UAE
Alnahhal, M. J., Alali, H., & Alshamsi, R. (2024) · International Journal of Emerging Technologies in Learning (iJET) · doi:10.3991/ijet.v19i06.49799 · Cited by 3
Reeping, D. P., & Shah, A. (2024) · ASEE Annual Conference & Exposition · doi:10.18260/1-2--47047 · Cited by 3
Ethical Considerations for Generative AI in Social Science Research
Brown, B. A., & Heitner, K. L. (2024) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3693-8557-9.ch006 · Cited by 3
Speaking with a “forked tongue” – misalignment between user ratings and textual emotions in LLMs
Yang, Y., & Huang, J. (2024) · Kybernetes · doi:10.1108/k-06-2024-1458 · Cited by 3
Alam, M. S., Asmawi, A., Haque, M. H., Patwary, M. N., Ullah, M. M., & Fatema, S. (2024) · Iraqi Journal for Computer Science and Mathematics · doi:10.52866/ijcsm.2024.05.03.016 · Cited by 6
“A qualitative study was designed to explore ChatGPT’s ongoing role in education [12, 18, 20, 29, 40, 49].”
Metadata
- Authors
- Rahman, M. M., & Watanobe, Y.
- Affiliations
- University of Aizu (JP)
- Corresponding author
- Rahman, Md. Mostafizer
- Venue
- Applied Sciences
- Publisher
- MDPI AG
- Published
- 2023-05-08
- Volume
- 13
- Issue
- 9
- Pages
- 5783
- ISSN
- 2076-3417 (ISSN-L)
- OpenAlex
- W4381982883
- Semantic Scholar
- 3b69ec0fc17c5134c9b002bebe283e0189bfdae5
- Cited by
- 1,080
- References
- 58
- Citations by year
- 2026: 193 · 2025: 338 · 2024: 425 · 2023: 124
- FWCI
- 253.56
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Applied Sciences (publishedVersion, cc-by)
- Preprints.org (acceptedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Japan Society for the Promotion of Science (JSPS) KAKENHI (23H03508)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Physical Sciences › Computer Science › Computer Science Applications › Teaching and Learning Programming
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- ChatGPT, large language models, artificial intelligence in education, academic dishonesty, critical thinking, programming education, code generation, code correction, student and teacher perceptions
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.
A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
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
