Bots and Beginners: AI- and PST-Generated Social Studies Lesson Plans

Summary

10 social studies teacher educators rated AI-generated lesson plans higher than their preservice teacher-generated counterparts, indicating there is a place for AI in the day-to-day work of teaching.

Abstract 

This study explored the comparative quality of social studies lesson plans generated by artificial intelligence (AI) and preservice teachers (PSTs), assessed by experienced social studies teacher educators. Using a concurrent mixed-methods approach, evaluators blind-reviewed six lesson plans—three AI-generated and three PST-generated—using a research-based rubric addressing coherence, clarity, learning objectives, activities, assessment alignment, and integration of inquiry and critical thinking. Quantitative findings indicated AI-generated lessons consistently outperformed PST-generated lessons in clarity, structured alignment, and inquiry integration. Qualitative analyses revealed a central paradox: AI was better at the technical elements of lesson planning, excelling in procedural efficiency and clarity, making them practically beneficial for immediate classroom use; however, human teachers brought more depth and creativity when those elements were present, demonstrating greater historical nuance and critical engagement. These findings suggest AI’s promising potential as a foundational resource to alleviate teacher workloads and reduce burnout, while underscoring the irreplaceable role of human creativity and pedagogical judgment.

Primary research question

How do AI- and PST-generated social studies lesson plans compare in quality when blind evaluated by social studies teacher educators?

What the research builds on

Many teacher preparation programs (TPPs) seek to develop pre-service teachers’ (PSTs) capacity for generating high-quality lesson plans. Some schools even require teachers turn formal lesson plans into their administration (Baeder,n.d.). As curricular-instructional gatekeepers, teachers should and do possess at least some level of control over how curricular standards are transformed into educational experiences in the classroom. This responsibility is perhaps even more acute in the context of social studies education, as social studies teachers carry the additional responsibility of deciding timeless issues tied to the social studies curriculum: whose history is taught, how, and to what end. In an era increasingly defined by fear and defensive teaching in response to so-called “divisive issues” legislation passed throughout the United States (Kaka et al., 2024), social studies teachers’ responsibility to design lessons characterized by inclusivity, equity, and the pursuit of justice has become even more crucial. At the same time, however, teachers have only so many hours in the workday, and lesson planning presents non-trivial demands on their cognitive and temporal resources, possibly detracting from other vital elements of their job.For instance, taking the time to construct whole lesson plans in the way taught in many TPPs may draw time and band-width away from providing timely, thoughtful feedback, a crucial part of quality, equitable teaching and learning.Electing to bring some of that work home, as many do, presents the added risk of teacher burnout. Machine learning tools, often referred to as artificial intelligence (AI), may hold promise for bootstrapping teachers’ pedagogical expertise to provide quality lesson plans while also cutting down on the cognitive and temporal demands of lesson planning, as well as mitigating constraints related to institutional resources like time and mate-rials, freeing teachers up to focus more on students and providing quality feedback.

What the research adds to the discussion

This study offers empirical evidence to help ground conversations emerging around the nascent topic of AI use in social studies education. Our findings provide important insights into how AI-generated lesson plans compare in quality to those created by PSTs at the end of their undergraduate preparation program. While others, such as Clark and van Kessel (2024), questioned the pedagogical utility of AI-generated lessons, the findings of this study complicate that position. We found that in nearly every dimension evaluated (coherence, clarity, alignment of objectives and assessments, and integration of inquiry), participants rated the AI-generated lesson plans higher than those created by PSTs. It is worth reminding readers that seasoned teacher educators provided these rating. These results suggest that at the very least, AI offers a strong starting point for lesson planning and may reduce some of the cognitive and temporal burdens that contribute to teacher burnout. We entered this work through a technoskeptical lens (Krutka & Carpenter, 2017; Selwyn, 2019), critically questioning the efficacy and appropriateness of integrating technological innovations such as generative AI into educational contexts. While we anticipated AI would be markedly constrained in its ability to replicate the relational, contextual, and human-centered practices central to social studies education, our data pushed us to a more nuanced framework that better reflects our findings. In line with the TPACK framework (Mishra & Koehler, 2006), our findings suggest AI holds particular affordances for the procedural aspects of teaching—such as organizing content, sequencing instruction, and aligning objectives with assessments—which PSTs still struggle to master. The AI-generated lessons demonstrated strengths in clarity and structure, offering practical, teachable materials many participants said they would prefer to use in their own classrooms. PST-created lessons were often more creative and contextually rich, especially in fostering historical nuance and critical thinking. While AI-generated lessons included inquiry elements, they were frequently described as surface-level or formulaic. The PST-created lessons, though inconsistently executed, occasionally achieved deeper engagement through primary source analysis and complex essential questions. This points to a central paradox: AI was better at the technical elements of lesson planning, but human teachers brought more depth and creativity when those elements were present. These findings raise interesting questions about how we use technology in education. Indeed, these questions high-light timeless issues in education through a contemporary lens. Like SmartBoards, projectors, Google Docs, and the many other technological innovations that have populated classrooms over the decades, every technology always carries both affordances and constraints (Koehler & Mishra, 2009; Mishra & Koehler, 2006). The key is to use technology for the things it does better and/or more efficiently than humans, and leave creative processes to human thinking and feeling. In the case of this study, AI proved better thanPSTs at the procedural work of lesson planning (e.g. follow-ing a lesson design sequence, keeping the lesson flow clear and organized, aligning objectives and assessment). When it came to the consequential work of lesson planning, however, the human PSTs outperformed AI (e.g. identifying and designing opportunities for higher-order thinking and deeper engagement). One reason for this maybe a lack of quality social studies lesson plans available online for AI platforms to metabolize into plans they generate in response to a prompt. Or, on the flip side, and possibly more the case, it may not be a lack of quality but rather a preponderance of the kind of low-quality lessons aligning to the kind of deskilled teaching bemoaned in the social studies education field for decades. In any case, this study offers another early jumping-off point for future research into how social studies education might integrate AI and other emerging technologies. This research underscores AI’s potential to significantly alleviate teacher workloads and address the increasingly urgent issue of teacher burnout (Steiner et al., 2022; Walker,2021). Given the cognitive and temporal demands associated with lesson planning, integrating AI could free teachers from routine planning tasks, thus allowing them to focus more intensively on relational, creative, and more pedagogically complex tasks. AI-generated plans could also effectively fulfill these administrative requirements for schools requiring teachers submit formalized lesson plans, thus reducing unnecessary cognitive and temporal burdens on teachers. Reallocating teacher attention toward aspects of teaching that uniquely benefit from human insight, such as formative assessment, differentiated instruction, culturally responsive teaching, and building strong teacher-student relationships, could significantly enhance educational out-comes and teacher satisfaction. Nevertheless, perhaps the most crucial consideration arising from our findings pertains to the implications for teacher education: Should PSTs continue investing significant instructional time in learning traditional lesson planning when AI demonstrates substantial competence in this domain? On the one hand, dedicating precious instructional resources to teach-ing a skill that AI can perform proficiently may seem redundant. On the other hand, engaging PSTs deeply in the lesson design process may hold intrinsic educational value, fostering critical pedagogical skills and deeper content knowledge. It is also important to recognize that the lesson plans created byPSTs in this study were generated at a specific point in their professional development trajectory. All were created and taught during student teaching, which was their final semester in their undergraduate TPP. Anecdotally, it is also well-documented that many experienced teachers rarely produce formal lesson plans in daily practice, opting instead for abbreviated plans or adaptive teaching strategies responsive to real-time classroom dynamics. Given these considerations, a potential pathway forward for TPPs could involve reorienting instruction to combine lesson-planning pedagogy with critical engagement withAI-generated resources. TeachingWorks (n.d.) emphasizes that successful novice teachers are those adept at selecting and adapting existing instructional materials rather than creating lessons entirely anew. Leveraging AI-generated lesson plans as “objects to think with” could facilitate PSTs’ pedagogical judgment, allowing them to critically analyze, refine, and adapt materials, thereby building deeper pedagogical insights and professional acumen. Moving forward,AI-generated lesson plans may serve as valuable objects to think with in TPP classrooms, allowing PSTs to engage in critique, modification, and thoughtful adaptation. In this way, AI becomes not a substitute for teacher judgment but a catalyst for its development.

Novel methodology 

The study leveraged Ph.D.-holding social studies teacher educators as expert evaluators to provide a rigorous, blinded expert review of the phenomena under study.

Implications for society 

While this study highlights the promise of AI to aid teachers in their day-to-day work, it also highlights a timeless tension: Whenever teachers get a firm handle on their work and become better and more efficient at it, schooling institutions are rarely content to leave them be, often preferring to fill any time teachers gain back through the use of technology with new tasks. We hope this will not be the case with AI, but generations of research on educational technology casts doubt on this prospect of teachers leveraging technology to achieve greater balance in their personal and professional lives.

Implications for research 

This study adds to the limited, nascent work being done on AI in education and in social studies education more specifically.