Why critical thinking and AI belong in the same conversation
School systems are under growing pressure to prepare students for a world shaped by rapid technological change. At the same time, educators are navigating persistent concerns about whether students are developing the critical thinking and problem-solving skills that employers, communities, and higher education increasingly expect. Artificial intelligence (AI) is often presented as a solution—but the reality inside many classrooms is more complicated.
A 2024 qualitative case study published in TOJET: The Turkish Online Journal of Educational Technology examined how middle school educators in Michigan view students’ critical thinking skills and how (or whether) teachers are using AI to support problem-solving and deeper learning. The findings are instructive not because they showcase a polished AI rollout, but because they reveal the practical barriers schools face when AI is discussed more than it is implemented.
Study snapshot: What the researchers examined
The study, led by Dr. Aimee Lintner, used a constructivist framework—an approach that emphasizes learning as an active process where students build new understanding through experience and prior knowledge. In a constructivist classroom, the teacher’s role is essential: guiding inquiry, structuring experiences, and helping students reflect and connect ideas.
To understand how AI fits into this reality, the research team conducted a qualitative case study in a Michigan public middle school serving grades 6–8. The school context matters: it was an urban Title I setting with significant linguistic diversity (over ninety languages represented in the district). Data collection included:
Classroom observations (approximately 30 minutes per classroom)
Teacher interviews (10 open-ended questions)
Review of lesson plan materials
Participants represented a range of teaching experience (roughly five to twenty years) and subject areas including math, science, language arts, STEM, and special education.
Key finding #1: AI use was minimal—and confidence was lower than interest
One of the clearest outcomes was that AI was generally not being used in classrooms, aside from two lessons that incorporated limited AI elements. Importantly, teachers were not broadly “anti-AI.” Many expressed curiosity and saw potential value, but they lacked a starting point and felt overwhelmed by the pace of change and day-to-day demands.
Teachers in the study described a gap between the expectation to innovate and the practical supports needed to do so. Although professional development (PD) requirements existed (e.g., minimum annual hours), educators reported they had not attended PD focused on AI implementation in education.
In market terms, this is a classic adoption barrier: interest exists, but usability, training, and workflow integration are insufficient. When AI tools feel like “one more initiative,” they are unlikely to move from experimentation to consistent classroom practice.
Key finding #2: Educators perceived critical thinking as a significant student need
Across interviews, teachers expressed concern that students were not progressing adequately in critical thinking and problem-solving. Several educators suggested that students are often “spoon-fed,” which reduces opportunities to struggle productively, justify reasoning, and learn from mistakes—core components of critical thinking.
Observations reinforced this theme. Instruction was frequently whole-group and teacher-led, followed by students practicing the same skill in the same manner. While some teachers attempted to prompt deeper thinking, students often needed more structured support to sustain academic conversation and expand their reasoning.
The implication is not that teachers are unwilling to teach critical thinking. Rather, it suggests that classroom structures, time constraints, and varying student readiness levels can make it difficult to consistently create the conditions where critical thinking develops.
Key finding #3: Diversity and personalization needs are rising faster than traditional instruction can accommodate
The study highlights a challenge familiar to many school leaders: classrooms contain a wide range of skill levels, background knowledge, and language needs. Teachers described students who were multiple grade levels behind alongside peers performing at or above grade level. In this context, “one lesson for everyone” can unintentionally widen gaps.
Several educators emphasized that students need more choice and personalization. However, the study found little evidence of AI-supported customization in practice. Personalization that did occur was limited (for example, students choosing writing ideas when summarizing), rather than adaptive pathways informed by ongoing data.
This is where AI is often positioned as a lever: not to replace teaching, but to help educators differentiate instruction at scale. Yet the study’s findings suggest that without targeted implementation support, AI remains more promise than practice.
What constructivism suggests about effective AI use
Because the study is grounded in constructivist theory, it implicitly offers a useful lens for evaluating AI tools. In a constructivist environment, AI should not merely accelerate worksheet completion or automate grading. Instead, it should help students engage in authentic tasks, reflect on their thinking, and connect learning to prior knowledge and real-world contexts.
Examples of constructivist-aligned AI use in middle school could include:
Guided inquiry prompts that help students generate hypotheses, test ideas, and revise conclusions
Scaffolded discussion supports that help multilingual learners participate in academic conversation
Project-based learning supports (planning, outlining, feedback cycles) that keep students in control of the work while receiving structured guidance
Adaptive practice that targets prerequisite gaps so students can access grade-level tasks more successfully
Notably, these uses require intentional instructional design. AI does not create critical thinkers by default; it can either deepen thinking or shortcut it depending on how tasks are structured.
Practical barriers schools must address before expecting results
The case study also points to a set of operational realities that school systems should treat as prerequisites, not afterthoughts:
Professional development that is specific, not generic: Teachers need concrete models, sample lesson structures, and time to practice.
Clear guidance on ethical and responsible use: Concerns about privacy, bias, and academic integrity must be addressed through policy and training.
Workflow fit: Tools should reduce complexity, not add to it. If AI requires extensive setup or troubleshooting, adoption will stall.
Instructional clarity: Teachers need clarity on the “why” (learning goals) before the “what” (tool features).
Where TinyEYE fits in the broader school readiness picture
Although this study focuses on classroom instruction, its conclusions intersect with a broader reality: students’ ability to engage in critical thinking is influenced by communication skills, language development, self-regulation, and confidence—areas that can be supported through specialized services.
As an online therapy provider serving schools, TinyEYE operates in the same ecosystem of constraints highlighted in the study: staffing pressures, diverse learner needs, and the demand for individualized support. When schools strengthen students’ foundational communication and learning readiness skills, they are better positioned to benefit from constructivist, student-centered instruction—whether AI-supported or not.
Actionable next steps for school and district leaders
Based on the study’s themes, schools considering AI to support critical thinking in middle grades can start with a phased, capacity-building approach:
Assess readiness: Identify current instructional patterns (whole-group vs. small-group), teacher comfort with technology, and student access.
Define “critical thinking” operationally: Choose observable behaviors (justifying answers, evaluating sources, revising work) and align tasks accordingly.
Invest in targeted PD: Provide training that includes classroom examples, prompt strategies, and safeguards—not just tool overviews.
Pilot with a small group: Start with a few teachers and one or two use cases (e.g., discussion scaffolds, inquiry prompts) and collect feedback.
Measure what matters: Look for changes in student reasoning, engagement, and quality of explanations—not only completion rates.
Conclusion: The opportunity is real, but implementation is the work
This case study underscores a central truth about AI in education: the gap is not only technological—it is organizational and instructional. Teachers in the study saw the need for stronger critical thinking and were open to AI’s potential, yet reported low confidence and minimal classroom use. In diverse middle school settings, personalization is increasingly necessary, but difficult to deliver consistently without strong supports.
AI may become a meaningful tool for constructivist, student-centered learning. However, schools will need clear goals, practical professional development, and responsible implementation plans to ensure AI strengthens—rather than substitutes for—students’ thinking.
For more information, please follow this link.
