pith. sign in

arxiv: 2605.02703 · v1 · submitted 2026-05-04 · 💻 cs.HC · cs.AI· cs.LG

ProPACT: A Proactive AI-Driven Adaptive Collaborative Tutor for Pair Programming

Anahita Golrang , Kshitij Sharma , Olga Viberg This is my paper

Pith reviewed 2026-05-08 18:53 UTC · model grok-4.3

classification 💻 cs.HC cs.AIcs.LG
keywords pair programmingcollaborative learningproactive feedbackadaptive tutoringjoint visual attentionmultimodal modelingAI in education
0
0 comments X

The pith

ProPACT uses forecasts of joint attention and effort to deliver proactive scaffolds that improve pair programming outcomes.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces ProPACT as a system that treats the pair itself as the learner rather than two separate individuals. It builds a model from joint visual attention, joint mental effort, and individual effort data to anticipate when collaboration is about to falter. An XGBoost model forecasts these states up to thirty seconds ahead so the tutor can step in with light guidance before performance drops. A controlled study with twenty-six programming pairs found measurable gains in debugging success, speed, and lasting improvements in how the pairs coordinated attention and effort. This matters for anyone designing tools that support real-time teamwork instead of isolated practice.

Core claim

ProPACT constructs a multimodal dyadic learner model based on Joint Visual Attention (JVA), Joint Mental Effort (JME), and individual mental effort, and employs an XGBoost-based forecasting model to predict emerging suboptimal collaboration states up to 30 seconds in advance. These predictions drive a hierarchical adaptive policy that delivers minimally intrusive scaffolds while fading support during productive collaboration. A within-subject study with 26 pair-programming dyads shows that proactive feedback significantly improves debugging success, task efficiency, feedback uptake, and post-intervention gains in JVA and JME.

What carries the argument

The XGBoost forecasting model operating on the multimodal dyadic learner model to predict suboptimal states 30 seconds ahead and trigger a fading scaffold policy.

If this is right

  • Proactive interventions raise debugging success rates compared with reactive or no feedback.
  • Task completion becomes more efficient when scaffolds arrive before coordination breaks down.
  • Learners show higher uptake of the feedback itself under the proactive schedule.
  • Gains in joint visual attention and joint mental effort persist after the intervention ends.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same forecasting-plus-fading approach could be tested in other real-time team tasks such as remote design reviews or surgical simulations.
  • If prediction windows can be extended or combined with additional sensors, the system might support larger groups without increasing intrusion.
  • The emphasis on fading support suggests a pathway for gradually transferring regulation skills from the AI to the human pairs.

Load-bearing premise

The multimodal dyadic learner model based on joint visual attention, joint mental effort, and individual mental effort, together with the XGBoost forecasting model, can accurately predict emerging suboptimal collaboration states up to 30 seconds in advance.

What would settle it

A follow-up experiment in which prediction accuracy falls below usable levels and the measured gains in debugging success, efficiency, and post-intervention JVA/JME disappear.

Figures

Figures reproduced from arXiv: 2605.02703 by Anahita Golrang, Kshitij Sharma, Olga Viberg.

Figure 1
Figure 1. Figure 1: Feedback Tools categorized as High, Average, or Low. These states are compared against a desired col￾laboration matrix and mapped to predefined trigger conditions, with feedback selected via a top-down hierarchical policy that prioritizes minimal intervention and esca￾lates support only when forecasted risk of breakdown increases. Trigger conditions are summarized in view at source ↗
Figure 2
Figure 2. Figure 2: Hybrid AI Framework view at source ↗
Figure 3
Figure 3. Figure 3: Debugging success , time, and uptake across the proactive feedback conditions (RQ1) 4 Discussion Building on previous systematic analysis of regulation in collaborative learning sys￾tems [23], which highlights the predominance of individual-level metrics over group and multi-level regulation, this study investigated the effectiveness of ProPACT, the AI-driven, proactive adaptive learning system that models… view at source ↗
Figure 4
Figure 4. Figure 4: Process Measurements (RQ2) that proactive multimodal feedback significantly improved pair-programming perfor￾mance: pairs supported by the adaptive system resolved more bugs and completed debugging tasks more efficiently than those in the control condition. These findings empirically support claims that learning regulation can be triggered by early cognitive or behavioral signals rather than only after ove… view at source ↗
read the original abstract

Effective pair programming depends on coordination of attention, cognitive effort, and joint regulation over time, yet most adaptive learning systems remain individual-centric and reactive. This paper introduces ProPACT, a proactive AI-driven adaptive collaborative tutor that treats collaboration itself as the object of instruction. ProPACT constructs a multimodal dyadic learner model based on Joint Visual Attention (JVA), Joint Mental Effort (JME), and individual mental effort, and employs an XGBoost-based forecasting model to predict emerging suboptimal collaboration states up to 30 seconds in advance. These predictions drive a hierarchical adaptive policy that delivers minimally intrusive scaffolds while fading support during productive collaboration. A within-subject study with 26 pair-programming dyads shows that proactive feedback significantly improves debugging success, task efficiency, feedback uptake, and post-intervention gains in JVA and JME, demonstrating the potential of forecast-driven dyadic adaptivity for real-time collaborative learning regulation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces ProPACT, a proactive AI-driven adaptive collaborative tutor for pair programming. It constructs a multimodal dyadic learner model using Joint Visual Attention (JVA), Joint Mental Effort (JME), and individual mental effort, then applies an XGBoost-based forecasting model to predict emerging suboptimal collaboration states up to 30 seconds ahead. These forecasts drive a hierarchical adaptive policy that delivers minimally intrusive scaffolds and fades support during productive periods. A within-subject study with 26 pair-programming dyads reports that proactive feedback yields significant gains in debugging success, task efficiency, feedback uptake, and post-intervention improvements in JVA and JME.

Significance. If the forecasting model is shown to be reliable, the work advances the field by demonstrating forecast-driven dyadic adaptivity rather than reactive individual scaffolding. The within-subject design with 26 dyads is a clear empirical strength, enabling direct comparison of proactive versus baseline conditions. The integration of multimodal signals (JVA/JME) into a real-time policy is a substantive contribution to collaborative learning systems.

major comments (2)
  1. The description of the XGBoost forecasting model (Section 4.2) reports no performance metrics whatsoever: no accuracy, AUC, precision-recall at varying lead times, cross-validation scheme, or comparison against a reactive baseline. This is load-bearing for the central claim that predictions enable effective proactive scaffolding; without these numbers the observed study gains could be produced by any feedback rather than by its forecast-driven timing.
  2. Results section (Section 5): the abstract and text assert statistically significant improvements in debugging success, efficiency, uptake, and JVA/JME gains, yet supply no p-values, effect sizes, confidence intervals, or details on data-handling rules (normality, multiple-comparison correction, outlier policy). These omissions prevent assessment of whether the within-subject findings robustly support the proactive-policy conclusion.
minor comments (2)
  1. Figure 3 (policy diagram) and the accompanying text use inconsistent terminology for 'suboptimal states' versus 'emerging suboptimal collaboration'; a single definition and cross-reference would improve clarity.
  2. The participant demographics and task description (Section 5.1) are brief; adding a table of dyad characteristics and exact task instructions would aid reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the thoughtful and constructive review. The comments identify key areas where additional detail will strengthen the manuscript's claims regarding the forecasting model and statistical reporting. We address each major comment below and will incorporate revisions as noted.

read point-by-point responses

  1. Referee: The description of the XGBoost forecasting model (Section 4.2) reports no performance metrics whatsoever: no accuracy, AUC, precision-recall at varying lead times, cross-validation scheme, or comparison against a reactive baseline. This is load-bearing for the central claim that predictions enable effective proactive scaffolding; without these numbers the observed study gains could be produced by any feedback rather than by its forecast-driven timing.

    Authors: We agree that detailed performance metrics for the XGBoost model are necessary to substantiate the proactive nature of the scaffolding. The current manuscript emphasizes the end-to-end system and empirical outcomes but does not include these evaluations. In the revised manuscript, we will add accuracy, AUC-ROC, precision-recall curves across lead times (0-30s), the cross-validation scheme (e.g., time-series split or dyad-level CV), and a direct comparison to a reactive baseline that triggers on current rather than forecasted states. This addition will demonstrate that the observed benefits derive from reliable forecast-driven timing rather than generic feedback. revision: yes

  2. Referee: Results section (Section 5): the abstract and text assert statistically significant improvements in debugging success, efficiency, uptake, and JVA/JME gains, yet supply no p-values, effect sizes, confidence intervals, or details on data-handling rules (normality, multiple-comparison correction, outlier policy). These omissions prevent assessment of whether the within-subject findings robustly support the proactive-policy conclusion.

    Authors: We acknowledge that the statistical reporting in Section 5 is incomplete. The within-subject design with 26 dyads used paired tests (t-tests or Wilcoxon signed-rank as determined by Shapiro-Wilk normality checks), with Bonferroni correction applied for the four primary outcomes. In the revision, we will report exact p-values, Cohen's d effect sizes, 95% confidence intervals for all key metrics, and explicit details on normality testing, outlier policy (e.g., >3 SD from mean), and multiple-comparison correction. This will enable readers to fully assess the robustness of the proactive-policy effects. revision: yes

Circularity Check

0 steps flagged

No significant circularity; claims rest on empirical study results

full rationale

The paper's core contribution is an empirical within-subject study with 26 dyads demonstrating gains in debugging success, efficiency, feedback uptake, and post-intervention JVA/JME improvements from proactive scaffolds. The multimodal dyadic model and XGBoost forecaster are presented as implemented components whose outputs drive the policy, but no derivation chain, equations, or self-citations reduce any central claim to a fitted parameter or self-definition. The absence of reported XGBoost validation metrics (accuracy, lead-time performance) is a separate evidence gap rather than circularity; the study results remain independent of any internal reduction to inputs.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The approach rests on standard machine-learning assumptions plus domain assumptions about attention and effort metrics; no new entities are introduced.

free parameters (1)
  • XGBoost hyperparameters and prediction thresholds
    The forecasting model requires fitting multiple parameters to collaboration data.
axioms (1)
  • domain assumption JVA, JME, and individual mental effort are valid proxies for dyadic collaboration quality
    The learner model is built directly on these signals.

pith-pipeline@v0.9.0 · 9225 in / 1119 out tokens · 67357 ms · 2026-05-08T18:53:32.793590+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

41 extracted references · 41 canonical work pages

  1. [1]

    99-106 (2023)

    Sharma, K., Molinari, G.: What happens when collaborators are not in-synch? In: Proceedings of the 16th International Conference on Computer-Supported Collaborative Learning-CSCL 2023, Pp. 99-106 (2023). International Society of the Learning Sciences

    work page 2023

  2. [2]

    In: Proceedings of the 2017 Acm Sigcse Technical Symposium on Computer Science Education, pp

    Rodríguez, F.J., Price, K.M., Boyer, K.E.: Exploring the pair programming process: Characteristics of effective collaboration. In: Proceedings of the 2017 Acm Sigcse Technical Symposium on Computer Science Education, pp. 507–512 (2017) 11

    work page 2017

  3. [3]

    British Journal of Educational Technology51(6), 2391–2406 (2020)

    Järvelä, S., Gašević, D., Seppänen, T., Pechenizkiy, M., Kirschner, P.A.: Bridging learning sciences, machine learning and affective computing for understanding cognition and affect in collaborative learning. British Journal of Educational Technology51(6), 2391–2406 (2020)

    work page 2020

  4. [4]

    International journal of computer-supported collaborative learning14(4), 425–441 (2019)

    Järvelä, S., Järvenoja, H., Malmberg, J.: Capturing the dynamic and cyclical nature of regulation: Methodological progress in understanding socially shared regulation in learning. International journal of computer-supported collaborative learning14(4), 425–441 (2019)

    work page 2019

  5. [5]

    Journal of Computer Assisted Learning39(5), 1714–1731 (2023)

    Liu, C.-Y., Li, W., Huang, J.-Y., Lei, L.-Y., Zhang, P.-R.: Collaborative pro- gramming based on social shared regulation: An approach to improving students’ programming achievements and group metacognition. Journal of Computer Assisted Learning39(5), 1714–1731 (2023)

    work page 2023

  6. [6]

    Frontline Learning Research8(2), 1–34 (2020)

    Bakhtiar, A., Hadwin, A.F.: Dynamic interplay between modes of regulation during motivationally challenging episodes in collaboration. Frontline Learning Research8(2), 1–34 (2020)

    work page 2020

  7. [7]

    In: Proceed- ings of the 15th International Conference on Computer-Supported Collaborative Learning-CSCL 2022, Pp

    Sharma,K., Olsen, J.K.:What bringsstudents together?: Investigating thecausal relationship between joint mental effort and joint visual attention. In: Proceed- ings of the 15th International Conference on Computer-Supported Collaborative Learning-CSCL 2022, Pp. 131-138 (2022). International Society of the Learning Sciences

    work page 2022

  8. [8]

    Educational technology research and development72(3), 1383–1418 (2024)

    Ke, F., Liu, R., Sokolikj, Z., Dahlstrom-Hakki, I., Israel, M.: Using eye-tracking in education: review of empirical research and technology. Educational technology research and development72(3), 1383–1418 (2024)

    work page 2024

  9. [9]

    In: 2021 IEEE Virtual Reality and 3D User Interfaces (VR), pp

    Bueno-Vesga, J.A., Xu, X., He, H.: The effects of cognitive load on engagement in a virtual reality learning environment. In: 2021 IEEE Virtual Reality and 3D User Interfaces (VR), pp. 645–652 (2021). IEEE

    work page 2021

  10. [10]

    polat et al

    Polat, H., Kayaduman, H., Taş, N., Battal, A., Kaban, A., Bayram, E.: Learning declarative and procedural knowledge through instructor-present videos: learn- ing effectiveness, mental effort, and visual attention allocation: H. polat et al. Educational technology research and development73(6), 3479–3513 (2025)

    work page 2025

  11. [11]

    International Society of the Learning Sciences, Proceedings

    Sharma, K., Olsen, J., Verma, H., Caballero, D., Jermann, P.: Challenging Joint Visual Attention as a Proxy for Collaborative Performance: ISLS Annual Meet- ing 2021 (virtual): International Society of the Learning Sciences. International Society of the Learning Sciences, Proceedings. 91–98 (2021) (2021)

    work page 2021

  12. [12]

    Interacting with Computers33(2), 115–133 (2021) 12

    D’angelo, S., Schneider, B.: Shared gaze visualizations in collaborative inter- actions: Past, present and future. Interacting with Computers33(2), 115–133 (2021) 12

    work page 2021

  13. [13]

    Sharma, K., Giannakos, M.: Multimodal data capabilities for learning: What can multimodal data tell us about learning? British Journal of Educational Technology51(5), 1450–1484 (2020)

    work page 2020

  14. [14]

    In: Proceedings of the Seventh International Learning Analytics & Knowledge Conference, pp

    Spikol, D., Prieto, L.P., Rodríguez-Triana, M.J., Worsley, M., Ochoa, X., Cukurova, M., Vogel, B., Ruffaldi, E., Ringtved, U.L.: Current and future multimodal learning analytics data challenges. In: Proceedings of the Seventh International Learning Analytics & Knowledge Conference, pp. 518–519 (2017)

    work page 2017

  15. [15]

    grey-box

    Sharma, K., Papamitsiou, Z., Giannakos, M.: Building pipelines for educational data using ai and multimodal analytics: A “grey-box” approach. British Journal of Educational Technology50(6), 3004–3031 (2019)

    work page 2019

  16. [16]

    Wohltjen, S., Wheatley, T.: Interpersonal eye-tracking reveals the dynamics of interacting minds. Front. Hum. Neurosci. 18 (2024)

    work page 2024

  17. [17]

    British Journal of Educational Technology55(5), 1900–1925 (2024)

    Yan,L.,Echeverria,V.,Jin,Y.,Fernandez-Nieto,G.,Zhao,L.,Li,X.,Alfredo,R., Swiecki, Z., Gašević, D., Martinez-Maldonado, R.: Evidence-based multimodal learning analytics for feedback and reflection in collaborative learning. British Journal of Educational Technology55(5), 1900–1925 (2024)

    work page 1900

  18. [18]

    British Journal of Educational Technology56(2), 712–733 (2025)

    Edwards, J., Nguyen, A., Lämsä, J., Sobocinski, M., Whitehead, R., Dang, B., Roberts, A.-S., Järvelä, S.: Human-ai collaboration: Designing artificial agents to facilitate socially shared regulation among learners. British Journal of Educational Technology56(2), 712–733 (2025)

    work page 2025

  19. [19]

    In: Proceedings of the 2015 10th Joint Meeting on Foundations of Software Engineering, pp

    Moreno, G.A., Cámara, J., Garlan, D., Schmerl, B.: Proactive self-adaptation under uncertainty: a probabilistic model checking approach. In: Proceedings of the 2015 10th Joint Meeting on Foundations of Software Engineering, pp. 1–12 (2015)

    work page 2015

  20. [20]

    Journal of Learning Analytics7(1), 59–71 (2020)

    Kent, C., Cukurova, M.: Investigating collaboration as a process with theory- driven learning analytics. Journal of Learning Analytics7(1), 59–71 (2020)

    work page 2020

  21. [21]

    Frontiers in Artificial Intelligence, 6, 1151391 (2023)

    Mallik, S., Gangopadhyay, A.: Proactive and reactive engagement of artificial intelligence methods for education: a review. Frontiers in Artificial Intelligence, 6, 1151391 (2023)

    work page 2023

  22. [22]

    Scientific Reports15(1), 39797 (2025)

    Alkan, B.B., Kuzucuk, S., Alkan, N., Sinan, A.: Using machine learning to pre- dict student outcomes for early intervention and formative assessment. Scientific Reports15(1), 39797 (2025)

    work page 2025

  23. [23]

    British Journal of Educational Technology55(4), 1398–1436 (2024) 13

    Sharma, K., Nguyen, A., Hong, Y.: Self-regulation and shared regulation in col- laborativelearninginadaptivedigitallearningenvironments:Asystematicreview of empirical studies. British Journal of Educational Technology55(4), 1398–1436 (2024) 13

    work page 2024

  24. [24]

    In: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, pp

    Duchowski, A.T., Krejtz, K., Biele, C., Niedzielska, A., Kiefer, P., Raubal, M., Giannopoulos, I.: The index of pupillary activity: Measuring cognitive load vis- à-vis task difficulty with pupil oscillation. In: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, pp. 1–13 (2018)

    work page 2018

  25. [25]

    International Journal of Computer-Supported Collaborative Learning 13(3), 241–261 (2018)

    Schneider, B., Sharma, K., Cuendet, S., Dillenbourg, P.: Leveraging mobile eye- trackers to capture joint visual attention in co-located collaborative learning groups. International Journal of Computer-Supported Collaborative Learning 13(3), 241–261 (2018)

    work page 2018

  26. [26]

    99–106 (2023)

    Sharma, K., Molinari, G.: What happens when collaborators are not in-synch? In: Proceedings of the 16th International Conference on Computer-Supported Collaborative Learning (CSCL 2023), pp. 99–106 (2023)

    work page 2023

  27. [27]

    Frontiers in psychology5, 510 (2014)

    Coco, M.I., Dale, R.: Cross-recurrence quantification analysis of categorical and continuous time series: an r package. Frontiers in psychology5, 510 (2014)

    work page 2014

  28. [28]

    Efficiency, and Processes in Brownfield Programming Tasks.(06 2025) (2025)

    Shihab, M., Hundhausen, C., Tariq, A., Haque, S., Qiao, Y., Mulanda, B.: The effects of github copilot on computing students’ programming effectiveness. Efficiency, and Processes in Brownfield Programming Tasks.(06 2025) (2025)

    work page 2025

  29. [29]

    In: 2025 5th International Conference on Artificial Intelligence and Education (ICAIE), pp

    Choi, W.C., Peng, J., Choi, I.C., Lei, H., Lam, L.C., Chang, C.I.: Improving young learners with copilot: The influence of large language models (llms) on cognitive load and self-efficacy in k-12 programming education. In: 2025 5th International Conference on Artificial Intelligence and Education (ICAIE), pp. 284–288 (2025). IEEE

    work page 2025

  30. [30]

    In: Proceedings of the ACM 2012 Conference on Computer Supported Cooperative Work, pp

    Jermann, P., Nüssli, M.-A.: Effects of sharing text selections on gaze cross- recurrence and interaction quality in a pair programming task. In: Proceedings of the ACM 2012 Conference on Computer Supported Cooperative Work, pp. 1125–1134 (2012)

    work page 2012

  31. [31]

    In: Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems, pp

    D’Angelo, S., Begel, A.: Improving communication between pair programmers using shared gaze awareness. In: Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems, pp. 6245–6290 (2017)

    work page 2017

  32. [32]

    International Journal of Computer-Supported Collaborative Learning15(4), 469–498 (2020)

    Hayashi, Y.: Gaze awareness and metacognitive suggestions by a pedagogical conversational agent: an experimental investigation on interventions to sup- port collaborative learning process and performance. International Journal of Computer-Supported Collaborative Learning15(4), 469–498 (2020)

    work page 2020

  33. [33]

    In: Proceedings of the 2020 ACM Conference on Innovation and Technology in Computer Science Education, pp

    Marwan,S.,Dombe,A.,Price,T.W.:Unproductivehelp-seekinginprogramming: What it is and how to address it. In: Proceedings of the 2020 ACM Conference on Innovation and Technology in Computer Science Education, pp. 54–60 (2020)

    work page 2020

  34. [34]

    PloS one12(8), 0183884 (2017)

    Xie, H., Wang, F., Hao, Y., Chen, J., An, J., Wang, Y., Liu, H.: The more total cognitive load is reduced by cues, the better retention and transfer of multimedia 14 learning: A meta-analysis and two meta-regression analyses. PloS one12(8), 0183884 (2017)

    work page 2017

  35. [35]

    British Journal of Educational Technology54(5), 1057–1076 (2023)

    Järvelä, S., Nguyen, A., Hadwin, A.: Human and artificial intelligence collabo- ration for socially shared regulation in learning. British Journal of Educational Technology54(5), 1057–1076 (2023)

    work page 2023

  36. [36]

    In: Proceedings of the 6th ACM Conference on Conversational User Interfaces, pp

    Edwards, J., Nguyen, A., Sobocinski, M., Lämsä, J., De Araujo, A., Dang, B., Whitehead, R., Roberts, A.-S., Kaarlela, M., Jarvela, S.: Mai-a proactive speech agent for metacognitive mediation in collaborative learning. In: Proceedings of the 6th ACM Conference on Conversational User Interfaces, pp. 1–5 (2024)

    work page 2024

  37. [37]

    In: Proceedings of the 16th International Conference of the Learning Sciences-ICLS 2022, Pp

    Nguyen, A., Järvelä, S., Wang, Y., Rosé, C.P.: Exploring socially shared regula- tion with an ai deep learning approach using multimodal data. In: Proceedings of the 16th International Conference of the Learning Sciences-ICLS 2022, Pp. 527-534 (2022). International Society of the Learning Sciences

    work page 2022

  38. [38]

    Education and Information Technologies, 1–39 (2025)

    Kim, J., Detrick, R., Yu, S., Song, Y., Bol, L., Li, N.: Socially shared regulation of learning and artificial intelligence: Opportunities to support socially shared regulation. Education and Information Technologies, 1–39 (2025)

    work page 2025

  39. [39]

    Assessment & Evaluation in Higher Education50(1), 127–143 (2025)

    Wood, J., Pitt, E.: Empowering agency through learner-orchestrated self- generated feedback. Assessment & Evaluation in Higher Education50(1), 127–143 (2025)

    work page 2025

  40. [40]

    Information16(7), 519 (2025)

    Li, M., Wilson, J.: Ai-integrated scaffolding to enhance agency and creativity in k-12 english language learners: A systematic review. Information16(7), 519 (2025)

    work page 2025

  41. [41]

    Electronic Journal of e-Learning23(4), 126–141 (2025) 15

    Rakhmetov,M.,Sadvakassova,A.,Saltanova,G.,Kuanbayeva,B.,Zhusupkalieva, G.: Evaluation of an ai-based feedback system for enhancing self-regulated learning in digital education platforms. Electronic Journal of e-Learning23(4), 126–141 (2025) 15

    work page 2025