research, not Research: seeing the Forest and the trees in one swell foop

The AI panic in education is anchored in a chronic institutional paralysis: teachers positioned as passive victims of forces beyond their control, students framed as deficient, and solutions imposed from above. When teachers investigate their own practice, however, they transform from anxious consumers of expert opinions into confident designers of learning. In the best scenarios, they collaborate with their colleagues.

They probe their own questions, collect their own evidence, and try out solutions grounded in what actually happens with their students. This shift from deficit thinking to inquiry thinking seems critical for removing the debris from an AI-disrupted world where the old playbooks have crumbled.

What a difference this perspective might make becomes clear when we examine how teacher-researchers would reframe the current AI crisis. Instead of lamenting what students lack or what technology threatens, they'd ask: What can I learn? What can I test? What can I change?

*****

Dixie Goswami and Peter Stillman co-edited what we used as our textbook in a year-long teacher research group circa 1988 mentored by Jim Hahn, an Area III writing project treasure from Northern California whom I wished I could have had in high school. "Reclaiming the Classroom: Teacher Research as an Agency for Change" provided our group with access to thought leaders in teacher research who helped to midwife a period I fondly recall as “Whole Language.”

We learned about the power of perspective that year, about a lens on the classroom now theorized pragmatically in the literature as “teacher noticing,” a shift that occurs when teachers inhabit the role of researcher-in-practice. As participant-observers, they are practitioners who double as theorists examining and articulating the underlying principles of their work. They talk less and listen more.

Why is this double duty important? It makes them begin to state their intentions ("I'm doing this experiment because I believe..."), to test their underlying assumptions ("Does this actually work the way I think it does?"), and to search for links between a mental model of teaching and what happens in the “crucible of the classroom,” a phrase I picked up from another of my mentors, David Pearson. Teacher-research did that for me.

This theorizing process is powerful because it's grounded in real classroom experience. I wrote two teacher research papers that year, each based on quasi-ethnographic methods. Conflicted about the most effective way to increase student ownership of their writing, giving them wide open choice or assignments as usual, I needed to find out what my fourth graders thought. I interviewed each one and took notes, slowing down to get them to restate what they just said so I could capture it verbatim. They loved it.

I learned these things. Some students thought assignments were easier because they didn’t have to work as hard, a sentiment that rings true across the grade levels, I’ve found. Some thought assignments were boring; they made you feel like “when you have to fold the clothes”—another cultural constant I’ve seen again and again. In the end, I gave them a choice. I’d give them an assignment if they wanted it or they could tell me what they were assigning themselves to do. The concept of ownership was more complex than I at first realized. I think the buzz word today is “agency.”

*****

When teachers articulate their intentions, they're forced to be deliberate about their choices. When they test assumptions, they discover some of their long-held beliefs about learning don't hold up well in the crucible. But they can take it further. When they look for connections between what they do and what Research “says,” they can bridge the gap between the observational visitations of external educational researchers and the concrete realities of teaching.

LLMs are a godsend for teacher researchers searching for bridges between capital R Research and practice, particularly AI search models like Perplexity and others (cf: Educating AI on Substack). Much useful Research can be accessed via open sources. As an example of AI use which extends rather than offloads human intelligence, a teacher prompting Perplexity to do a deep search about current research on the longitudinal relationship between pure explicit, systematic phonics instruction in first grade and critical comprehension in high school, say, big R papers, is apt. Check it out. What quantitative evidence supports the claim that phonics in first grade causes critical comprehension in high school?

*****

This shift in perspectives transforms teachers from consumers of research to research producers. During objective observations and measurements during visitation research, the researcher maintains distance, trying to minimize their influence on what they're studying. The mind operates in a detached, analytical mode, e.g., categorizing, quantifying, looking for patterns that might be replicated and generalized. There's a clear subject-object distinction.

But participant-observation erases that distance. The teacher-researcher is the observer and the observed, the watcher and the watched, taking up a social metacognitive stance that I found thrilling during that year with Jim Hahn while I taught fourth grade. I designed the learning environment and experienced its effects. But I emulated the visiting researcher by documenting and reflecting on what I saw and heard in Steno notebooks, document collection, and audio recordings. This activity makes for a different kind of knowing, an embodied, relational, and phenomenological sense of knowledge. I learned what it means to know something in your bones.

Teacher-researchers learn to understand how knowledge is constructed, what makes evidence credible, and how context shapes findings. This inside knowledge makes them much more critical readers of others' work. Less apt to accept uncritically others' theories reflects a retreat from deference to an embrace of discernment. Teacher-researchers who can evaluate methodology without its mystique, question the assumptions that render the method sound, and assess whether the findings are relevant to their specific context, are less vulnerable to coercion, binary thinking, or fads.

Teacher-researchers can ask harder questions. What evidence supports this? How was this tested? Does this align with what I'm observing in my own classroom? More authoritative in their assessment of papers, they develop intellectual confidence grounded in their own systematic inquiry. They're not dealing in opinions or preferences or anecdotes about curricula and methods. They're making informed judgments based on evidence they've gathered and analyzed themselves.

How might these quotes from the frenzy around cheating and AI stirred up recently by the New York Times be rewritten from the pens of teacher-researchers? (The final exercise comes from a different source.) Your task? Read the quote. Take a moment to write down your version as a teacher-researcher. Then have a look at my version (don’t cheat or I’ll have to make a phone call home).

Exercise 1

From the article: "Cheating by copying from A.I. is rampant, particularly among my disengaged seniors who are simply biding their time until graduation."

Your version:

From a fictional teacher-researcher: "I've been deliberately tracking my seniors' writing this semester and noticed more AI-generated work, especially from students I'd already documented as disengaged. I'm wondering what aspects of my assignments or grading might be pushing them toward this shortcut, and how I could redesign things to get them actually invested in their own thinking again."

Exercise 2

"Eighth graders accepted that as their answer without questioning its validity... They simply lack the foundational knowledge or the intellectual resilience to challenge such unlikely claims."

Your version:

From a fictional teacher-researcher: "My eighth graders took that answer at face value, which got me thinking about what I'm seeing in their critical thinking development. I'm wondering if it's about not having enough background knowledge to spot the red flags, or if there's something about how I'm structuring discussions that doesn't encourage them to push back on ideas. I want to try some different approaches and see what happens when I give them more scaffolding for questioning sources."

Exercise 3

"The duration her students are willing to engage with challenging material has significantly decreased... if a student struggles to grasp something within a few minutes, they are more inclined to abandon their own cognitive efforts in favor of seeking help from a chatbot or a peer."

Your version:

From a fictional teacher-researcher: "I've been tracking how long my students stick with difficult problems, and I'm seeing too many of them give up faster than they used to. When they hit a wall after just a couple minutes, they're turning to ChatGPT or asking classmates instead of working through it themselves. I'm curious whether this is about the difficulty level of what I'm assigning, how I'm teaching them to approach struggle, or if there's something about the classroom environment that makes asking for help feel easier than persisting. I want to experiment with different ways of supporting productive struggle."

Exercise 4

From the Hilarious Bookbinder, who published a post on Substack on March 25, 2025, with this conclusion. While the Bookbinder’s frustration and sadness about the current situation in higher education paints a dismal picture, as I recall, the post doesn’t assign responsibility solely to AI:

Hilarious Bookbinder: "One thing all faculty have to learn is that the students are not us. We can't expect them all to burn with the sacred fire we have for our disciplines... Our job is to kindle that flame, and we're trying to get that spark to catch, but it is getting harder and harder and we don't know what to do."

Your version:

From a fictional faculty-researcher: "I've been reflecting on my assumption that students should naturally share my passion for this subject, and I'm realizing that's not realistic or fair. I'm seeing my job as finding ways to help them discover their own connections to the material, but my usual approaches aren't working like they used to. I want to document what I'm trying and track what actually sparks genuine interest. Maybe I need to study how my students engage with things they do care about and see if I can bridge from there."

Three Discussion Questions

1. The Shift from Deficit to Design How might educational institutions support teachers in making this shift from deficit thinking ("students lack X") to design thinking ("what can I change about my approach")? What systemic barriers prevent teachers from adopting this researcher stance, and what would need to change to make teacher research more than just an individual choice?

2. AI as Research Tool vs. Research Subject You mention LLMs as a "godsend for teacher researchers" when used to extend rather than offload human intelligence. How do we help teachers distinguish between using AI to enhance their inquiry capabilities versus studying their questions about AI impact on learning with AI? What does it look like when teachers research with AI rather than just research about AI?

3. From Consumer to Producer The transformation from research consumer to research producer seems crucial for teachers navigating the current AI disruption. But how do we scale this beyond individual teacher-researchers? The most impactful teacher-research is done in communities. What would it mean for entire schools or districts to adopt this inquiry stance toward the challenges they're facing with technology and student engagement?