The Role of Artificial Intelligence in Green Education: Optimizing Teacher Workflow and Enhancing Pedagogical Design under Sustainable Development Pedagogy (SDP) Constraints

Synthesis: Talebzadeh (2026) conducts a quasi-experimental study with 28 pre-service teacher teams, finding that AI-assisted Sustainable Development Pedagogy constraints significantly improve instructional design quality (t(27) = 13.78, p < 0.001, Cohen's d = 2.80). The intervention transformed teachers from conventional designers into strategic educational managers.

Key Findings

• 28 pre-service teacher teams compared across Baseline and SDP phases using the Integrated AI Triad (IAT) model
• Statistically significant improvement in overall design quality: t(27) = 13.78, p < 0.001
• Exceptionally large effect size: Cohen's d = 2.80
• SDP workflow compliance showed the largest gain (+1.83 units)
• Zero-paper resource management constraint acted as a catalyst for more reflective, strategic instructional design
• Teacher role transformation: from conventional pedagogical designer to efficient, reflective educational manager

What this means for practice

• Instructors. Replace open-ended AI access with a hard design constraint — this study used a zero-paper, fully digitalized assessment workflow — because the constrained phase produced the study's largest gain, in SDP workflow compliance (+1.83 units, d = 3.50); the authors read that as structured Scaffolding succeeding where unstructured AI use does not.
• Instructors. Re-engineer an existing lesson plan under the new constraint rather than writing a fresh one, mirroring the baseline → SDP sequence in which the same teams resubmitted their design.
• Faculty developers. Train teams to write constraint-specific prompts (the SDP prompt set) alongside the general IAT prompts, so the AI supports the required workflow instead of routing around it.
• Faculty developers. Score submissions on all four rubric indices — overall design score, SDP workflow compliance, strategic reflection, and complexity and innovation — so the reflection the constraint forces becomes visible and discussable.
• Faculty developers. Hold identical correction windows and final deadlines for both submissions and keep the two-week gap between phases that this study used, to limit practice-order and instructor-feedback confounds.

Limitations

• The study rests on 28 teams drawn from 91 pre-service geography teachers in one Geography Teaching Methods course at one university (Farhangiyan University, Tehran), so it is a single-site, single-discipline result.
• The design is pretest–posttest within the same teams and includes no separate control group: the same 28 teams produced both the baseline and the constrained plans, with concurrent instructor feedback on both sets serving as the paper's stated control rather than an independent comparison.
• The outcome is rubric-scored lesson-plan quality over a two-week window, not student learning; the rubric's complexity-and-innovation index measures higher-order cognitive demands embedded in the designed tasks, so the IAT model's emphasis on student cognitive outcomes goes unmeasured.

Citation

Talebzadeh, H. (2026). The Role of Artificial Intelligence in Green Education: Optimizing Teacher Workflow and Enhancing Pedagogical Design under Sustainable Development Pedagogy (SDP) Constraints. EdArXiv. doi:10.35542/osf.io/x6qzy_v1.