The DOI matches its Crossref record; 34 of the 34 references it cites were found in a registry or an index; 7 of the 28 sources its claims cite were checked for support.
AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting
Open accessPublished version of a preprint
Abstract
Advances in generative artificial intelligence show great potential for improving education. Yet little is known about how this new technology should be used and how effective it can be compared to current best practices. Here we report a randomized, controlled trial measuring college students' learning and their perceptions when content is presented through an AI-powered tutor compared with an active learning class. The novel design of the custom AI tutor is informed by the same pedagogical best practices as employed in the in-class lessons. We find that students learn significantly more in less time when using the AI tutor, compared with the in-class active learning. They also feel more engaged and more motivated. These findings offer empirical evidence for the efficacy of a widely accessible AI-powered pedagogy in significantly enhancing learning outcomes, presenting a compelling case for its broad adoption in learning environments.
Overview
CoChat Summary
TL;DR
In a randomized crossover trial in an undergraduate physics course, students learned more from a custom AI tutor than from an in-class active learning lesson on identical content. The AI condition also produced higher engagement and motivation, while taking less time on task.
Key findings
- Students in the AI-tutored group had higher median post-test scores than the in-class active learning group: 4.5 (N = 142) vs 3.5 (N = 174), compared with a pre-test baseline median of 2.75 (N = 316).
- The post-test difference between AI tutoring and in-class active learning was statistically significant: z = -5.6, p < 10–8.
- In the linear regression model controlling for pre-test score, midterm exam, Force Concept Inventory, prior ChatGPT experience, topic, test version, and time on task, the AI group still performed substantially better; the paper reports p < 10–8 and an effect size estimate of 0.63, with a quantile-regression estimate of 0.73 to 1.3 standard deviations.
- Students in the AI group reported higher engagement (4.1 vs 3.6; t(311) = -4.5, p < 0.0001) and motivation (3.4 vs 3.1; t(311) = -3.4, p < 0.001); enjoyment and growth mindset were not significantly different.
- For time on task, the AI group had a median of 49 minutes; 70% spent less than 60 minutes and 30% more than 60 minutes. The paper reports no correlation between time on task and post-test scores.
Methods
- Randomized, controlled crossover design in Harvard’s introductory physics course (PS2) during Fall 2023; each of the 194 eligible students experienced both conditions across two lessons in consecutive weeks.
- Two lesson topics were used: surface tension and fluid flow; students took pre-tests and post-tests around each lesson, and the control and experimental conditions used identical content/worksheets, differing only in delivery (in-class active learning vs AI tutor “PS2 Pal”).
- Main analysis compared post-test scores using a two-sample rank-sum (Mann–Whitney) test and a linear regression model with clustering at the student level and controls for prior knowledge, prior course performance, FCI, ChatGPT experience, topic, test version, and time on task.
Limitations
- The paper states that the AI tutor’s success may depend on several contextual factors: a heterogeneous student population, high-quality instructional videos, a GPT-4-capable model, expert-crafted prompts, a carefully structured scaffold, and content that lends itself to this format.
- The authors explicitly say they do not presume structured AI tutoring will always outperform in-class active learning in all contexts, especially those requiring complex synthesis of multiple concepts and higher-order critical thinking.
- Evident from the text: the study was conducted in a single course at one institution and on two physics topics, so generalizability beyond that setting is not directly tested here.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures



Verification
Claims the paper makes
3 Valid · 2 Not valid · 11 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
“Although passive lectures are among the least effective modes of instruction, they remain in wide use in science, technology, engineering, and mathematics (STEM) courses6–8.”
§ AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting > Introduction With their human-like conversational style and knowledge drawn from extremely large data sets, generative artificial intelligence (GAI) chatbots have inspired visions of expert tutors available on demand through every smartphone1. The President of the United States, at the time of this investigation in 2023, pledged to “shape AI’s potential to transform education by creating resources to support educators deploying AI-enabled educational tools, such as personalized tutoring in schools.”1 Despite this recent excitement, previous studies show mixed results on the effectiveness of learning, even with the most advanced AI models2,3. While these models can answer technical questions, their unguided use lets students complete assignments without engaging in critical thinking. After all, AI chatbots are generally designed to be helpful, not to promote learning. They are not trained to follow pedagogical best practices (e.g., facilitating active learning, managing cognitive load4,1 and promoting a growth mindset).2 Another well-known flaw of AI tutors is their uncanny confidence when giving an incorrect answer or when marking a correct reply as incorrect5.3 As reported here, a carefully designed AI tutoring system, using the best current GAI technology and deployed appropriately, can not only overcome these challenges but also address significant known pedagogical challenges in an accessible way that can offer world-class education to any community or learning environment with an internet connection. Although passive lectures are among the least effective modes of instruction, they remain in wide use in science, technology, engineering, and mathematics (STEM) courses6–8. Passive lectures have several long-known issues: 1. They move too quickly for some students and too slowly for others because the teacher controls the pace of instruction; 2. Students do not receive personalized feedback to their questions as they arise; and 3. They fail to maintain consistent student engagement. Active learning pedagogies,4 such as peer instruction, small-group activities, or a flipped classroom structure, have demonstrated significant improvements over passive lectures9–14. However, any approach that involves one teacher working with many students will suffer, at least in part, from the same three problems that plague passive lectures. Working with an expert personal tutor is generally regarded as the most efficient form of education15. A tutor can guide the student while providing personalized feedback and answering questions as they arise. Expert tutors will adapt their approach to a student’s individual ability, pace, and specific needs. They offer a more focused and efficient learning experience, managing the student’s cognitive load. In addition, personalized instruction can foster a growth mindset, which has been shown to promote students’ persistence in the face of difficulties16,17. While the advantages of personalized instruction are clear, this model of education cannot scale to meet the needs of a large number of students15. What if an AI tutor could mimic the learning experience one would get from an expert (human) tutor? It could address the unique needs of each individual through timely feedback while adopting what is known from the science of how students learn best. This is the focus of our work. Through a design that involves targeted, content-rich prompt engineering, we developed an online tutor that uses GAI and best practices from pedagogy and educational psychology to promote learning in undergraduate science education. We conducted a randomized controlled experiment in a large undergraduate physics course (N = 194) at Harvard University, with a student population broadly representative of those found across a range of institutions, to measure the difference between 1) how much students learn and 2) students’ perceptions of the learning experience when identical material is presented through an AI tutor compared with an active learning classroom.
What the sources say
- Partly supportedcites [6] Henderson, C. & Dancy, M. H. Barriers to the use of research-based instructional strategies: The influence of both individual and situational characteristics. Phys. Rev. Spec. Top.-Phys. Educ. Res. 3(2), 020102 (2007). sourceExperimental — accuracy not yet measured
The Introduction (page unavailable) says instructors are encouraged to adopt research-based strategies because students will learn more, and the Practices section classifies one-sided discourse with passive students as traditional instruction. The paper does not rank passive lectures among the least effective modes of instruction. (LLM, conf=0.82)
- Partly supportedcites [7] Stains, M. et al. Anatomy of STEM teaching in North American universities. Science 359(6383), 1468–1470 (2018). sourceExperimental — accuracy not yet measured
The paper’s concluding discussion (page unavailable) describes lecturing and other teacher-centered practices as having limited impact. That supports the claim’s general point, but the paper does not rank passive lectures among the least effective modes of instruction. (LLM, conf=0.84)
- Not checked — no full text of the sourcecites [8] Handelsman, J. et al. Scientific teaching. Science 304, 521–522 (2004). 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.
- Partly supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“These previous approaches did not adhere to the same research-based best practices that informed our approach.”
§ AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting > Discussion > Designing successful student-AI interactions
What the sources say
- Not matchedcites Krupp et al. (2023)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Could not verifycites Forero (2023) sourceExperimental — accuracy not yet measured
Our copy of Forero & Herrera-Suárez (2023) contains no assessment of whether previous studies or approaches on AI-powered instruction adhered to research-based best practices. The nearest statement is in the Introduction (Section I): 'there are very few studies available to assess how student learning and performance are affected by the use of new AI-based technology like ChatGPT [25]' — a claim of scarcity, not of methodological deficiency, and the paper also only lists differing published positions [20]-[23] (including a news report on the NYC ChatGPT ban) without critiquing their practices. Likewise, Materials and Methods (Section II) describes its own design (same lecturer, identical topics/sequence, conventional pedagogy plus ChatGPT 3.5 introduction) but never frames it as informed by 'research-based best practices'. Searched: 'did not adhere to research-based best practices', 'limitations of prior research / methodological shortcomings', 'prior literature lacking rigorous experimental design', 'pedagogical principles / evidence-based instructional design'. (LLM, conf=0.68)
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.
- Not matched
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“They are not trained to follow pedagogical best practices (e.g., facilitating active learning, managing cognitive load4,1 and promoting a growth mindset).2”
§ AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting > Introduction With their human-like conversational style and knowledge drawn from extremely large data sets, generative artificial intelligence (GAI) chatbots have inspired visions of expert tutors available on demand through every smartphone1. The President of the United States, at the time of this investigation in 2023, pledged to “shape AI’s potential to transform education by creating resources to support educators deploying AI-enabled educational tools, such as personalized tutoring in schools.”1 Despite this recent excitement, previous studies show mixed results on the effectiveness of learning, even with the most advanced AI models2,3. While these models can answer technical questions, their unguided use lets students complete assignments without engaging in critical thinking. After all, AI chatbots are generally designed to be helpful, not to promote learning. They are not trained to follow pedagogical best practices (e.g., facilitating active learning, managing cognitive load4,1 and promoting a growth mindset).2 Another well-known flaw of AI tutors is their uncanny confidence when giving an incorrect answer or when marking a correct reply as incorrect5.3 As reported here, a carefully designed AI tutoring system, using the best current GAI technology and deployed appropriately, can not only overcome these challenges but also address significant known pedagogical challenges in an accessible way that can offer world-class education to any community or learning environment with an internet connection. Although passive lectures are among the least effective modes of instruction, they remain in wide use in science, technology, engineering, and mathematics (STEM) courses6–8. Passive lectures have several long-known issues: 1. They move too quickly for some students and too slowly for others because the teacher controls the pace of instruction; 2. Students do not receive personalized feedback to their questions as they arise; and 3. They fail to maintain consistent student engagement. Active learning pedagogies,4 such as peer instruction, small-group activities, or a flipped classroom structure, have demonstrated significant improvements over passive lectures9–14. However, any approach that involves one teacher working with many students will suffer, at least in part, from the same three problems that plague passive lectures. Working with an expert personal tutor is generally regarded as the most efficient form of education15. A tutor can guide the student while providing personalized feedback and answering questions as they arise. Expert tutors will adapt their approach to a student’s individual ability, pace, and specific needs. They offer a more focused and efficient learning experience, managing the student’s cognitive load. In addition, personalized instruction can foster a growth mindset, which has been shown to promote students’ persistence in the face of difficulties16,17. While the advantages of personalized instruction are clear, this model of education cannot scale to meet the needs of a large number of students15. What if an AI tutor could mimic the learning experience one would get from an expert (human) tutor? It could address the unique needs of each individual through timely feedback while adopting what is known from the science of how students learn best. This is the focus of our work. Through a design that involves targeted, content-rich prompt engineering, we developed an online tutor that uses GAI and best practices from pedagogy and educational psychology to promote learning in undergraduate science education. We conducted a randomized controlled experiment in a large undergraduate physics course (N = 194) at Harvard University, with a student population broadly representative of those found across a range of institutions, to measure the difference between 1) how much students learn and 2) students’ perceptions of the learning experience when identical material is presented through an AI tutor compared with an active learning classroom.
What the sources say
- Not checked — no full text of the sourcecites [4] Sweller, J. Cognitive load theory. In The Psychology of Learning and Motivation: Cognition in Education (eds Mestre, J. P. & Ross, B. H.) 37–76 (Elsevier Academic Press, 2011). source
- Not matchedcites [1] Singer, N. Will chatbots teach your children? New York Times https://www.nytimes.com/2024/01/11/technology/ai-chatbots-khan-education-tutoring.html (2024).
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Could not verifycites [2] Forero, M. G. & Herrera-Suárez, H. J. ChatGPT in the classroom: boon or bane for physics students’ academic performance? Preprint at arXiv:2312.02422 (2023). sourceExperimental — accuracy not yet measured
Our copy of Forero & Herrera-Suárez (2023) never addresses how chatbots were trained with respect to pedagogy. Section I only describes ChatGPT as 'a language model based on GPT (Generative Pre-trained Transformer) that performs conversational tasks' trained 'to generate relevant and coherent text sequences based on a prompt'; the Discussion (Sec. IV) limits itself to student-survey outcomes (reduced critical/independent thinking in ~30%, dissatisfaction with accuracy/explanations in >10%) and to the general points that such tools 'are not replacements for teachers' and should be used reflectively. Searched: 'not trained to follow pedagogical best practices', cognitive load / active learning / growth mindset, 'not designed for educational purposes', training objective — no passage states or contradicts the claim, so it cannot be judged from this source. (LLM, conf=0.70)
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.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Active learning pedagogies,4 such as peer instruction, small-group activities, or a flipped classroom structure, have demonstrated significant improvements over passive lectures9–14.”
§ AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting > Introduction With their human-like conversational style and knowledge drawn from extremely large data sets, generative artificial intelligence (GAI) chatbots have inspired visions of expert tutors available on demand through every smartphone1. The President of the United States, at the time of this investigation in 2023, pledged to “shape AI’s potential to transform education by creating resources to support educators deploying AI-enabled educational tools, such as personalized tutoring in schools.”1 Despite this recent excitement, previous studies show mixed results on the effectiveness of learning, even with the most advanced AI models2,3. While these models can answer technical questions, their unguided use lets students complete assignments without engaging in critical thinking. After all, AI chatbots are generally designed to be helpful, not to promote learning. They are not trained to follow pedagogical best practices (e.g., facilitating active learning, managing cognitive load4,1 and promoting a growth mindset).2 Another well-known flaw of AI tutors is their uncanny confidence when giving an incorrect answer or when marking a correct reply as incorrect5.3 As reported here, a carefully designed AI tutoring system, using the best current GAI technology and deployed appropriately, can not only overcome these challenges but also address significant known pedagogical challenges in an accessible way that can offer world-class education to any community or learning environment with an internet connection. Although passive lectures are among the least effective modes of instruction, they remain in wide use in science, technology, engineering, and mathematics (STEM) courses6–8. Passive lectures have several long-known issues: 1. They move too quickly for some students and too slowly for others because the teacher controls the pace of instruction; 2. Students do not receive personalized feedback to their questions as they arise; and 3. They fail to maintain consistent student engagement. Active learning pedagogies,4 such as peer instruction, small-group activities, or a flipped classroom structure, have demonstrated significant improvements over passive lectures9–14. However, any approach that involves one teacher working with many students will suffer, at least in part, from the same three problems that plague passive lectures. Working with an expert personal tutor is generally regarded as the most efficient form of education15. A tutor can guide the student while providing personalized feedback and answering questions as they arise. Expert tutors will adapt their approach to a student’s individual ability, pace, and specific needs. They offer a more focused and efficient learning experience, managing the student’s cognitive load. In addition, personalized instruction can foster a growth mindset, which has been shown to promote students’ persistence in the face of difficulties16,17. While the advantages of personalized instruction are clear, this model of education cannot scale to meet the needs of a large number of students15. What if an AI tutor could mimic the learning experience one would get from an expert (human) tutor? It could address the unique needs of each individual through timely feedback while adopting what is known from the science of how students learn best. This is the focus of our work. Through a design that involves targeted, content-rich prompt engineering, we developed an online tutor that uses GAI and best practices from pedagogy and educational psychology to promote learning in undergraduate science education. We conducted a randomized controlled experiment in a large undergraduate physics course (N = 194) at Harvard University, with a student population broadly representative of those found across a range of institutions, to measure the difference between 1) how much students learn and 2) students’ perceptions of the learning experience when identical material is presented through an AI tutor compared with an active learning classroom.
What the sources say
- Not checked — no full text of the sourcecites [9] Hake, R. R. Interactive-engagement vs. traditional methods: A six-thousand-student survey of mechanics test data for introductory physics courses. Am. J. Phys. 66, 64–74 (1998). source
- Not checked — no full text of the sourcecites [10] Crouch, C. H. & Mazur, E. Peer instruction: Ten years of experience and results. Am. J. Phys. 69, 970–977 (2001). source
- Not checked — no full text of the sourcecites [11] Deslauriers, L., Schelew, E., & Wieman, C. Improved learning in a large-enrollment physics class. Science 332, 862–864 (2011). source
- Not checked — no full text of the sourcecites [12] Freeman, S. et al. Active learning increases student performance in science, engineering, and mathematics. Proc. Natl. Acad. Sci. U.S.A. 111, 8410–8415 (2014). source
- Supportedcites [13] Fraser, J. M. et al. Teaching and physics education research: Bridging the gap. Rep. Prog. Phys. 77, 032401 (2014). source
The Introduction (p. 1) states plainly that 'the evidence for significantly higher conceptual understanding gains in active learning environments over passive ones is hard to overlook', and that 'the bulk of this research indicates that students taught interactively have considerably better conceptual understanding and problem-solving skills than students taught with lecture'. The same section quantifies it via Hake's (1998) meta-analysis (normalized gain 0.48 ± 0.14 for 48 interactive-engagement classes vs 0.23 ± 0.04 for 14 traditional courses) and Crouch & Mazur's Peer Instruction comparison, which covers the citing sentence's exemplars (peer instruction, small-group/hands-on activities). The 'flipped classroom' example is the citing authors' own illustrative list rather than something attributed here, and the comparative-improvement claim itself is what this citation supports. (LLM, conf=0.90)
- Partly supportedcites [14] Thornton, R. K., Kuhl, D., Cummings, K. & Marx, J. Comparing the force and motion conceptual evaluation and the force concept inventory. Phys. Rev. ST Phys. Educ. Res. 5, 010105 (2009). sourceExperimental — accuracy not yet measured
Thornton et al. (2009) do report a large advantage for research-based instruction over traditional lecture-type teaching: Table I in Sec. III B (Normalized gains on the FMCE) gives 63±6% average normalized gain for 2494 students in research-based classes versus 15±3% for 926 students in traditional classes across ten institutions, and Sec. IV (Final Thoughts and Conclusions) restates this as 'about 15%' versus 'about 60%'. However, the paper frames the contrast as 'traditional instruction' vs 'research-based methods or materials' in introductory college physics mechanics, assessed with the FMCE, and the table is offered as a reference range rather than a tested comparison of pedagogies — it reports no significance statistics for 'improvements over passive lectures'. The specific examples the claim carries are also only partly covered: Sec. III B lists cooperative-group problem solving, Interactive Lecture Demonstrations, peer instruction, tutorials, RealTime Physics, studio physics and workshop physics, but nothing on a flipped-classroom structure, which the source does not address. (LLM, conf=0.72)
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.
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 median learning gains for students, relative to the pre-test baseline (M = 2.75, N = 316), in the AI-tutored group were over double5 those for students in the in-class active learning group.”
§ AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting > Results > Learning gains: post-test scores
plainstates a causen = 316- 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.
18 more claims
Create a free account to see all 59 claims and to verify the 3 references still open.
- Probably not a substantial claim. 26
- It may state two separate claims as one. 14
- The quote only partly states this claim. 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 cited source says what the claim attributes to it.
In the citing paper
“while some studies suggest genAI can offer effective instruction [10-13]”
From AI tutoring can safely and effectively support students: An exploratory RCT in UK classrooms (2025) · cites this paper as [11]
The cited source says what the claim attributes to it.
In the citing paper
“Third, in many designs students choose how much time to spend on the task (e.g., Kazemitabaar et al., 2023; Kestin et al., 2025; Kreijkes et al., 2026), so the estimated effects combine the two margins: time allocation and learning productivity.”
From Experimental Evidence on the Learning Impact of Generative AI (2026) · cites this paper as Kestin et al., 2025
The cited source says what the claim attributes to it.
In the citing paper
“It was further demonstrated by Kestin et al. (2025) that AI pedagogically designed to require active pedagogical work outperformed conventional instructional contrasts on learning outcomes.”
From Effects of Partial vs. Full Cognitive Offloading with Generative AI on Learning Retention (2026) · cites this paper as Kestin et al., 2025
Statistics recomputed
Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.
- Consistentt(311) = -4.5reported p < .0001 · recomputed p < .001
“t(311) = − 4.5, p < .0001”
- Consistentt(311) = -3.4reported p < .001 · recomputed p < .001
“t(311) = -3.4, p < .001”
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Self-check results appear here once an independent check has confirmed them.
Citations
Cites
School Engagement: Potential of the Concept, State of the Evidence
Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004) · Review of Educational Research · doi:10.3102/00346543074001059 · Cited by 12,581 · cited 1×
“Key practices include (i) facilitating active learning 11,19 , (ii) managing cognitive load 4 , (iii) promoting a growth mindset 15,16 , (iv) scaffolding content 20 , (v) ensuring accuracy of information and feedback, (vi) delivering such feedback and information in a targeted and timely fashion 21…”
Active learning increases student performance in science, engineering, and mathematics
Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H., & Wenderoth, M. P. (2014) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1319030111 · Cited by 9,787
THE ROLE OF TUTORING IN PROBLEM SOLVING *
Wood, D., Bruner, J. S., & Ross, G. (1976) · Journal of Child Psychology and Psychiatry · doi:10.1111/j.1469-7610.1976.tb00381.x · Cited by 8,851 · cited 1×
“…include (i) facilitating active learning 11,19 , (ii) managing cognitive load 4 , (iii) promoting a growth mindset 15,16 , (iv) scaffolding content 20 , (v) ensuring accuracy of information and feedback, (vi) delivering such feedback and information in a targeted and timely fashion 21 , and (vii)…”
Mindset: The New Psychology of Success
(2006) · Choice Reviews Online · doi:10.5860/choice.44-2397 · Cited by 7,804
Hake, R. R. (1998) · American Journal of Physics · doi:10.1119/1.18809 · Cited by 6,072 · cited 1×
“Typical pre-instruction FCI scores in educational institutions range from aproximately 30% to 50% 30 .”
Sweller, J., Ayres, P., & Kalyuga, S. (2011) · doi:10.1007/978-1-4419-8126-4 · Cited by 5,531
Sweller, J. (2011) · Psychology of Learning and Motivation · doi:10.1016/b978-0-12-387691-1.00002-8 · Cited by 4,856
Shute, V. J. (2008) · Review of Educational Research · doi:10.3102/0034654307313795 · Cited by 4,631
Sweller, J. (2012) · Encyclopedia of the Sciences of Learning · doi:10.1007/978-1-4419-1428-6_446 · Cited by 4,405
Hestenes, D., Wells, M., & Swackhamer, G. (1992) · The Physics Teacher · doi:10.1119/1.2343497 · Cited by 3,235
Research works printed but not linked: 2
- Hake, R. R. Interactive-engagement versus traditional methods: A six-thousand-student survey of mechanics test data for introductory physics courses. Am. J. Phys. 66, 64–74 (1998). · DOI · Paper page
- Freeman, S. et al. Active learning increases student performance in science, engineering, and mathematics. Proc. Natl. Acad. Sci. 111(23), 8410–8415 (2014). · DOI · Paper page
Cited by
Nasr, N. R., Tu, C.-H., Werner, J., Bauer, T., Yen, C.-J., & Sujo-Montes, L. (2025) · Education Sciences · doi:10.3390/educsci15091198 · Cited by 75
“When thoughtfully integrated, as suggested by prior research, ChatGPT can become a catalyst for analysis and synthesis (Wang & Fan, 2025); without scaffolds, it risks becoming a shortcut around deep critical thinking (Kestin et al., 2025; Larson et al., 2024).”
Tromp, J. G., Schofield, D., Parvari, P. R., Poyade, M., Eaglesham, C., Torres, J. C., Johnson, T., Jürivete, T., Lauer, N., Reyes‐Lecuona, A., et al. (2025) · Applied Sciences · doi:10.3390/app15147973 · Cited by 37
Omirali, A., Kozhakhmet, K., & Zhumaliyeva, R. (2025) · Sustainability · doi:10.3390/su17177567 · Cited by 25
Generative AI in Self-Directed Learning: a thematic scoping review
Roe, J., & Perkins, M. (2025) · Interactive Learning Environments · doi:10.1080/10494820.2025.2603431 · Cited by 67
Melanou, C., & Beege, M. (2026) · Education Sciences · doi:10.3390/educsci16040651 · Cited by 12
Metadata
- Authors
- Kestin, G., Miller, K., Klales, A., Milbourne, T., & Ponti, G.
- Affiliations
- Harvard University (US)
- Corresponding author
- Kestin, Greg
- Venue
- Scientific Reports
- Published
- 2025-06-03
- Volume
- 15
- Issue
- 1
- Pages
- 17458-17458
- ISSN
- 2045-2322 (ISSN-L)
- PMC
- PMC12179260
- OpenAlex
- W4411505840
- Semantic Scholar
- 23c1bcb0c0450d79abbe0a1c2a9b4a3b60b6fe03
- Cited by
- 214
- References
- 31
- Citations by year
- 2026: 162 · 2025: 51
- FWCI
- 84.76
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by-nc-nd
- creativecommons.org/licenses/by-nc-nd/4.0 (tdm)
- creativecommons.org/licenses/by-nc-nd/4.0 (vor)
- Locations
- Scientific Reports (publishedVersion, cc-by-nc-nd)
- Research Square (acceptedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Harvard University
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
- Social Sciences › Education › Artificial Intelligence in Education
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- AI tutoring, generative AI in education, active learning, university students, learning outcomes, student engagement, student motivation, personalized learning
- MeSH terms
- Adult, Artificial Intelligence, Female, Humans, Learning, Male, Students, Universities, Problem-Based Learning, Young Adult
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 and qwen3.8-omni-flash).
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: AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting. https://app.cochat.ai/papers/10.1038/s41598-025-97652-6
- The paper
- Kestin, G., Miller, K., Klales, A., Milbourne, T., & Ponti, G. (2025). AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting. Scientific Reports. https://doi.org/10.1038/s41598-025-97652-6
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.