If I had a reliable, replicable program that created psychological safety during teacher observations, I would probably be writing books instead of blog posts.
I do not have that. Nobody does.
What I have is two years of research, ten honest conversations with teachers across the country, and a growing conviction that we are often asking the wrong question. We keep looking for the right feedback structure — the right sandwich, the right ratio of glow to grow, the right language for the debrief conversation. And those things matter. But they are not the answer to the problem most of us are actually trying to solve.
The problem is not the observation. The problem is the culture in which the observation happens.
When a teacher sees a principal's name on the walkthrough schedule, something happens. Before the day arrives, before anyone walks through the door, a calculation gets made — not consciously, but immediately. The culture of the building has already told the teacher how that day is going to go.
The question underneath the calculation is simple: Is it safe to be seen here?
Not seen performing. Seen actually teaching — imperfect, mid-process, still figuring things out. Safe to let someone watch the real classroom rather than the version prepared for company. Can I answer the knock in my jammies, or do I need to run upstairs and get dressed because I know I am being judged the moment I swing the door open?
One teacher I interviewed put it this way. He described navigating every conversation at work by running a mental calculation — "picking my words very carefully" — tracking who knew what about him and adjusting accordingly. "You got to keep a map, a flow chart of 'Is this person cool? What do they know about me?'"
He was not talking about observations specifically. He was talking about every interaction. Which is exactly the point.
Teachers follow principals on social media who seem to genuinely value and protect their people. They read those posts carefully. They compare them to what they experience on Monday morning. What is it like when the cameras are not rolling?
The teachers I interviewed were not describing observation protocols when they talked about feeling safe at work. They were describing something harder to reduce to a checklist and harder to fake.
Patrick, a veteran elementary teacher, described his principal Lisa in terms that stuck with me. "Every time she's in my room, she looks for the good," he said. "She finds every opportunity to look for the good and tell me what's happening from that reference point, from that frame. And I really appreciate her." She knows his husband. She has attended his dance performances in the community. She sees him — the whole person — and that is the lens she brings into the classroom.
That is not a feedback protocol. That is a relationship, built over time, that makes the observation possible.
The contrast is just as instructive. I spoke with teachers who described schools where inclusion looked fine on the surface — the right symbols, the right language, the right stated values — but where, when push came to shove, the parent's complaint carried more weight than the teacher's professional judgment. In those buildings, teachers knew it. They had seen it play out. And when the principal walked in for an observation, they performed accordingly.
Another teacher, Amanda, described sending an article she had written to one of her administrators — a piece on a topic relevant to a conversation they had been having. "He read the whole thing and came in and talked to me about it and was like, thank you so much," she said. Then she added, almost as an aside: "Do I think the others would all do that? Not necessarily. But the fact that there's anyone there who does is nice."
That last sentence is worth some extra attention. One administrator who treated her as a professional thinker, not just an employee to be managed, was enough to shift her experience of the building. That is how much this matters. And that is how little it takes to start.
I want to be honest about what I am not saying.
I am not saying feedback frameworks are useless. I am not saying the glow/grow conversation does not matter. I am not offering a solution, because every school is different — different history, different community, different people — and those differences make any universal fix a fiction.
What I am saying is that the outcomes we care about — teacher retention, student achievement, the discipline and attendance numbers that show up in the data — are downstream of something more fundamental than the observation process. They are downstream of whether teachers feel they can be themselves in the building where they work. Building that culture is slow. It is less measurable than a feedback rubric. It does not show up on any evaluation form. But it is what teachers are looking for, in every interaction, every small moment where a principal either backs their people or does not.
The observation is just the moment when everyone finds out what has already been built.
Charles Keller is an instructional coach and holds an EdD from Youngstown State University, where his dissertation examined psychological safety and identity development among K-12 educators. He is based in Phoenixville, Pennsylvania.
The technology is moving fast. The question of what it's actually for is moving slower.
Instructional leadership doesn't require a fresh start in August—it requires a system that helps principals consistently get into classrooms, provide meaningful feedback, and keep instruction at the center of their leadership.
New labor data draws a sharp line between AI that replaces a task and AI that assists with one. That line should be showing up in our lesson plans.
The instructional evidence platform for K–12 districts.
