arxiv: 2605.12934 · v1 · submitted 2026-05-13 · 💻 cs.CY

Recognition: 2 theorem links

· Lean Theorem

An Activity-Theoretical Approach to Teacher Professional Development in Pedagogical AI Agent Design

Haiyang Xin , Qiannan Niu , Shuang Li , Yimeng Sun , Ching Sing Chai , Lingyun Huang , Gaowei Chen

Authors on Pith no claims yet

Pith reviewed 2026-05-14 18:46 UTC · model grok-4.3

classification 💻 cs.CY

keywords teacher professional developmentpedagogical AI agentsCultural-Historical Activity TheorySelf-Determination Theorysystemic contradictionsteacher engagementformative intervention

0 comments

The pith

Systemic contradictions in teacher training programs cause 87% of educators to stop creating pedagogical AI agents within three weeks, even after completing the full professional development.

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

The study investigates why teachers abandon AI agent creation shortly after training despite acquiring the necessary skills. In the first cycle with 218 participants, behavioral data and interviews traced the rapid drop-off to contradictions within the activity system that blocked teachers' sense of autonomy, competence, and connection. A second cycle with 26 teachers tested a redesign grounded in Cultural-Historical Activity Theory and Self-Determination Theory that directly addressed those contradictions. The redesign produced higher sustained creation rates and simultaneous gains in both skill and motivation. The work reframes disengagement as a predictable response to misaligned systems rather than a teacher deficit and supplies a diagnostic method for redesigning future programs.

Core claim

In a two-cycle formative intervention, researchers identified systemic contradictions in standard teacher professional development as the primary driver of psychological need frustration, leading 87 percent of participants to cease AI agent creation within three weeks. Applying a CHAT-SDT redesign that resolved those contradictions produced synchronized improvements in both teachers' capacity to design agents and their willingness to continue doing so.

What carries the argument

The CHAT-SDT diagnostic framework, which uses Cultural-Historical Activity Theory to map contradictions in the activity system and Self-Determination Theory to link those contradictions to frustration of autonomy, competence, and relatedness needs.

If this is right

Redesigning professional development to remove identified contradictions raises both teachers' technical capacity and their sustained motivation to create AI agents.
Implementation failures in educational technology adoption can be treated as rational responses to need-thwarting activity systems rather than as evidence of teacher shortcomings.
The CHAT-SDT method supplies a replicable diagnostic sequence for diagnosing and correcting engagement problems in other forms of teacher professional development.
Simultaneous gains in capacity and willingness occur when contradictions are resolved at the system level instead of through additional skill training alone.

Where Pith is reading between the lines

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

The same diagnostic approach could be tested on professional development for other classroom technologies such as learning analytics dashboards or adaptive tutoring systems.
Policy makers could require activity-system audits before scaling new AI tools in schools to reduce the risk of rapid teacher disengagement.
Longer-term follow-up studies could check whether the redesigned programs produce measurable changes in classroom practice beyond the three-week window examined here.

Load-bearing premise

The improvements observed in the second cycle were caused by the targeted CHAT-SDT redesign rather than by smaller group size or other unmeasured differences between the two cycles.

What would settle it

A randomized controlled trial in which teachers assigned to the CHAT-SDT redesigned program show no higher three-week retention rate in agent creation than teachers in standard professional development would falsify the central claim.

read the original abstract

This two-cycle formative intervention study examined why teachers disengage from AI agent creation after professional development - a low engagement paradox - and tested whether systemic redesign could address it. Cycle 1 (N=218) revealed that despite completing comprehensive TPD, 87 percent of teachers ceased creating within three weeks, with behavioral tracking and interview analysis identifying systemic contradictions as the source of psychological need frustration rather than capacity deficits. Cycle 2 (N=26) implemented Cultural-Historical Activity Theory and Self-Determination Theory - driven redesign directly targeting diagnosed contradictions, achieving synchronized enhancement of both capacity and willingness. The findings reframe implementation failure as a rational response to need-thwarting systems and offer a replicable CHAT - SDT diagnostic framework for transformative professional development.

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.

Desk Editor's Note private letter to a colleague

The paper documents a real 87% dropout after teacher PD on AI agents and applies CHAT-SDT to blame systemic contradictions, but cycle 2's gains rest on a small uncontrolled sample.

read the letter

The main finding is straightforward: after a full professional development program, 87% of 218 teachers stopped creating pedagogical AI agents within three weeks. Behavioral logs and interviews point to contradictions in the activity system—such as mismatched tools, rules, and community expectations—frustrating autonomy and competence rather than any simple skill gap. Cycle 2 then redesigns the setup using Cultural-Historical Activity Theory and Self-Determination Theory on 26 teachers and reports better capacity and willingness outcomes. That descriptive pattern from cycle 1 is useful and grounded in actual tracking data. The CHAT-SDT diagnostic lens is a direct, if unsurprising, extension of existing theory to the AI-agent creation task. The redesign itself shows the authors trying to close the loop from diagnosis to intervention, which is more than many PD studies manage. The weakness is in cycle 2. The sample drops sharply from 218 to 26 with no randomization, no matched control arm, and no reported statistical tests or baseline equivalence checks. Smaller groups or selection of more motivated participants could explain the gains just as easily as the targeted contradiction fixes. Without those controls the causal claim stays provisional. This work is aimed at researchers and designers focused on teacher professional development in educational technology. Anyone running similar programs will recognize the dropout problem and may find the diagnostic framework worth testing. It deserves peer review because the initial empirical observation is clear and the theoretical move is coherent, but the methods section will need tighter controls and transparency before the intervention results can be taken as strong evidence.

Referee Report

1 major / 0 minor

Summary. This two-cycle formative intervention study uses Cultural-Historical Activity Theory (CHAT) and Self-Determination Theory (SDT) to diagnose and address teacher disengagement from pedagogical AI agent creation after professional development. Cycle 1 (N=218) documents an 87% dropout rate within three weeks despite completed training, attributing it via behavioral tracking and interviews to systemic contradictions causing psychological need frustration rather than capacity deficits. Cycle 2 (N=26) applies a targeted redesign of those contradictions, reporting synchronized gains in both capacity and willingness to create AI agents, and proposes a replicable CHAT-SDT diagnostic framework for transformative PD.

Significance. If the causal attribution holds, the work provides a concrete diagnostic framework for reframing implementation failures in educational AI as responses to need-thwarting activity systems rather than individual deficits, with potential to guide more effective teacher PD designs in technology integration.

major comments (1)

[Cycle 2] Cycle 2 (N=26): the claim that the CHAT-SDT redesign produced the observed synchronized capacity-willingness gains is undermined by the absence of controls, randomization, baseline comparisons, or statistical tests; the reduction from the Cycle 1 sample of 218 to a smaller non-randomized subset leaves selection effects and group-size differences as plausible alternative explanations for the improvement.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for highlighting the methodological limitations of Cycle 2. As a formative intervention study grounded in CHAT, the design prioritizes iterative diagnosis and redesign over experimental controls, but we agree the absence of randomization and statistical testing requires explicit discussion. We have revised the manuscript accordingly.

read point-by-point responses

Referee: [Cycle 2] Cycle 2 (N=26): the claim that the CHAT-SDT redesign produced the observed synchronized capacity-willingness gains is undermined by the absence of controls, randomization, baseline comparisons, or statistical tests; the reduction from the Cycle 1 sample of 218 to a smaller non-randomized subset leaves selection effects and group-size differences as plausible alternative explanations for the improvement.

Authors: We acknowledge that Cycle 2 was conducted without a control group, randomization, or formal statistical tests, as the study follows the principles of formative intervention research in CHAT, which emphasizes collaborative redesign of activity systems rather than hypothesis testing via RCTs. The N=26 subset consisted of teachers who voluntarily continued into the redesign phase after Cycle 1, introducing potential selection effects that we cannot rule out. However, the observed gains are tied to specific, pre-identified contradictions from Cycle 1 through repeated behavioral logs, pre/post activity system mappings, and interview data showing resolution of need-thwarting elements (e.g., division of labor and rules). We do not claim strict causal proof but rather evidence of synchronized improvement following targeted redesign. We will add a dedicated limitations subsection in the Discussion to explicitly address the lack of controls, non-random sampling, and interpretive nature of the findings, while strengthening the description of qualitative evidence supporting the gains. revision: partial

Circularity Check

0 steps flagged

No circularity; empirical claims rest on independent participant data and established external theories

full rationale

The paper reports a two-cycle formative intervention using direct behavioral tracking and interview data from N=218 (Cycle 1) and N=26 (Cycle 2) to measure disengagement rates and redesign effects. No equations, fitted parameters, or self-derived predictions appear; CHAT and SDT function as pre-existing external frameworks applied to the observations rather than being redefined or derived from the study's own outputs. Central inferences about contradictions and need frustration are grounded in the collected evidence and do not reduce to input definitions or self-citation chains. The analysis is therefore self-contained against external benchmarks with no load-bearing circular steps.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The study applies two established theories without introducing new mathematical parameters, invented entities, or ad-hoc axioms beyond standard domain assumptions of CHAT and SDT.

axioms (1)

domain assumption Cultural-Historical Activity Theory and Self-Determination Theory provide valid lenses for diagnosing contradictions and psychological need frustration in professional development systems
Invoked to interpret behavioral tracking and interview data as evidence of systemic rather than individual deficits

pith-pipeline@v0.9.0 · 5443 in / 1261 out tokens · 64041 ms · 2026-05-14T18:46:49.399581+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

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

IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

Cycle 1 revealed systemic contradictions (Tool-Subject, Rule-Object, Subject-Community) thwarting SDT needs; Cycle 2 redesign targeted need satisfaction to recouple capacity and willingness.
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

CHAT contradiction analysis and SDT need-frustration mapping

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

8 extracted references · 3 canonical work pages

[1]

An Activity-Theoretical Approach to Teacher Professional Development in Pedagogical AI Agent Design Haiyang Xin, Qiannan Niu, Shuang Li, Yimeng Sun Tony@cocorobo.cc, niuqiannan@cocorobo.cc, lishuang@cocorobo.cc, sunyimeng@cocorobo.cc COCOROBO Limited Ching Sing Chai, The Chinese University of Hong Kong, CSChai@cuhk.edu.hk Lingyun Huang, The Education Univ...

2024
[2]

between old structures and new objects. An expansive transformation occurs when the object and motive are reconceptualized to embrace radically wider possibilities, representing a collective journey through the zone of proximal development (Engeström, 2001). Table 1 illustrates how teachers’ AI agent creation activity system manifests across these six ele...

2001
[3]

capacity

Table 1 Activity System Components in Teachers’ AI Agent Creation Components Description in This Study Subjects In-service K-12 teachers participating in professional development workshops Object → Outcome Object: Addressing instructional challenges and student learning needs through AI-enhanced pedagogical approaches (TPACK) Outcome: Teachers achieving p...

2020
[4]

All surveys used 7-point Likert scales with strong reliability (α=0.88-0.94)

was administered post-only to assess whether the redesigned workshop successfully created a need-supportive environment. All surveys used 7-point Likert scales with strong reliability (α=0.88-0.94). Interview transcripts were independently coded by two researchers (Cohen's κ=0.78). Figure 2 CocoFlow AI Agent Builder: Editing Interface The editing interfac...

2023
[5]

Cycle 2 demonstrated that SDT-driven activity system redesign simultaneously shaped both capacity and willingness—rather than improving one and expecting the other to follow

that assumes training activities automatically translate into sustained practice. Cycle 2 demonstrated that SDT-driven activity system redesign simultaneously shaped both capacity and willingness—rather than improving one and expecting the other to follow. Three design principles emerged. First, diagnose systemic contradictions rather than surface barrier...

2001
[6]

For designers, choice-based task portfolios, discipline-specific scaffolded templates, and low-barrier peer sharing spaces operationalize need-supportive PD concretely. For researchers, the CHAT-SDT framework offers diagnostic tools to trace why specific adoptions succeed or fail—moving from generic barrier attribution toward context-specific contradictio...

work page doi:10.1191/1478088706qp063oa 2006
[7]

https://doi.org/10.1007/s43681-021-00074-z Chen, B

AI and Ethics, 2(1), 157-165. https://doi.org/10.1007/s43681-021-00074-z Chen, B. (2025). Beyond tools: Generative AI as epistemic infrastructure in education. arXiv preprint arXiv:2504.06928. Chen, S., Metoyer, R., Le, K., Acunin, A., Molnar, I., Ambrose, A., ... & Metoyer, R. (2025, July). Bridging the AI Adoption Gap: Designing an Interactive Pedagogic...

work page doi:10.1007/s43681-021-00074-z 2025
[8]

M., & Deci, E

https://doi.org/10.3390/su16073034 Ryan, R. M., & Deci, E. L. (2020). Intrinsic and extrinsic motivation from a self-determination theory perspective: Definitions, theory, practices, and future directions. Contemporary educational psychology, 61, 101860. Sikström, P., Valentini, C., Sivunen, A., & Kärkkäinen, T. (2024). Pedagogical agents communicating an...

work page doi:10.3390/su16073034 2020