Building data literacy together, one question at a time

Professional Learning Communities have a credibility problem in many schools. Not because the concept is wrong — the research on collaborative data use is strong — but because the implementation so often isn’t. Cult of Pedagogy documented it plainly: teachers describe their PLCs as “another box admin must check off,” consisting of “weekly meetings during our planning period that we have no say in and no benefit from.”

The gap between what PLCs are supposed to do and what they actually do is well-documented. A year-long case study of a US elementary school found that formally scheduled grade-level meetings “did not always produce the intended results” — not because teachers didn’t care, but because data literacy was assumed rather than built. Teams were expected to analyze together before they had a shared language for doing so.

Last week's post was about preparing individually for data meetings. This one is about what happens when you bring that preparation into a room with colleagues. Data literacy is a team muscle, and like any muscle, it develops through deliberate, repeated practice.

– The Shift

The shift isn’t from bad data meetings to good data meetings. It’s from data analysis as a meeting activity to data literacy as a team capacity. Those are different things.

A meeting produces an action plan. A data-literate team produces better questions and better action plans as a result.

Online there are consistently two points on what makes collaborative data use work. First, offering a safe environment for PLC members to ask questions is foundational. Without psychological safety, teachers perform certainty instead of demonstrating curiosity, and nothing gets learned. Second, collaboration only develops data literacy when the meeting has structure. Without a shared protocol, the most assertive voice tends to set the direction, and the team’s collective intelligence goes unused.

The protocol below is designed to build both: structure that slows the rush to action, and norms that make honest inquiry feel safe.

One more thing before the protocol. Data literacy at the team level requires explicit attention to what data should and shouldn’t travel. Student data carries legal protections under FERPA. It also carries ethical obligations that go beyond the law. What the team looks at together, what gets documented, what enters any digital tool — these decisions matter and need to be made deliberately, not by default.

– The Framework

The protocol builds directly on the three questions introduced in our previous post, the Data Meeting Survival Guide: What is this measuring? Who benefits? What does it ask us to do?

STEP 1 OBSERVE (10 min)
Each person looks at the same data source independently and in silence. Write only what you see. “Thirteen out of twenty-two students missed the inference question.” Not: “They haven’t been reading enough.” After silent observation, each person shares one thing they noticed without explaining it.
→ Why this step matters: Most teams skip this step entirely. They jump straight to interpretation because silence feels unproductive. But Argyris’s ladder of inference shows exactly what happens when we skip it: we run straight from raw data to conclusion, bypassing the reasoning that makes conclusions trustworthy. Ten minutes of shared observation changes the quality of everything that follows.

STEP 2 INTERPRET (10 min)
Now apply the three questions as a team. What is this data actually measuring and what might it be missing? What would explain the pattern we observed? Crucially: what are at least two different explanations, not just the first one that comes to mind? The goal is to generate multiple interpretations before the team commits to any one of them.
→ Why this step matters: Teams that generate only one explanation tend to act on it regardless of whether it’s right. Forcing a second and third interpretation slows the rush to solutions and surfaces the assumptions embedded in the first one. Different teachers often bring different explanations.

STEP 3 QUESTION (10 min)
From the interpretations on the table, identify the one that the team most wants to investigate further. Then ask: what would we need to see to know if that interpretation is right? What could we each try in our classrooms this week that would produce evidence one way or the other? These are small experiments, not action plans.
→ Why this step matters: The distinction between an experiment and an action plan is important. An action plan commits the team to a solution. An experiment tests a hypothesis. Experiments are less threatening, more honest, and — when they work — produce more durable changes in practice because they’re grounded in evidence from your actual students.

STEP 4 ACT (10 min)
Each teacher commits to one specific, small experiment before the next meeting. Not “I’ll differentiate more.” Something like: “I’m going to try a think-aloud with five students on Thursday and look at whether the gap is in inference or in background knowledge.” The commitment has a who, a what, a with whom, and a when. Document it.
→ Why this step matters: Small commitments get kept. Large commitments get deferred. The goal of this step isn’t an impressive action plan for the record. It’s a testable experiment that produces new data for the next meeting. A cycle of inquiry that improves with each pass is what builds a data-literate team over time.

FERPA	Before any discussion: agree as a team on what data stays in the room. Student names and individual identifying information should never be shared in whole-group discussion or entered into AI tools. Patterns and aggregated observations are fair game. When in doubt, err toward protecting student privacy.

– The Honest Part

Psychological safety is the prerequisite for all of this — and it’s the thing most often missing. Edmondson’s research is clear: when people don’t feel safe saying “I don’t know” or “I might be wrong,” they perform certainty instead of demonstrating curiosity. A PLC can have the best protocol in the world and produce nothing if the norms don’t support honest inquiry.

There are two specific failure modes that can impact the application of this protocol:

1. One or two teachers dominate the Interpret step. The protocol addresses this partly by requiring each person to share independently before discussion begins. If it persists, name it: “Before we move on, I want to hear from someone who hasn’t spoken yet.”
2. Admin presence changes the dynamic. When a principal or instructional coach is in the room, teachers often shift from genuine inquiry to performance. If that’s happening in your PLC, it’s worth a direct conversation with whoever facilitates: the meeting serves teachers best when it’s a place for honest questions, not evaluated answers.

And on the FERPA question: individual teacher data literacy won’t fix systemic data protection failures. But it will mean your team doesn’t contribute to them. That matters.

– Your Move

Collaboration Hub Action Item(s)
This week	Don’t propose this to your whole PLC yet. Find one colleague. Share one data question. Try the Observe step together: five minutes, the same data, independent notes, then share what you each noticed.
Next meeting	Bring the four-step protocol and propose it as an experiment for one meeting. Frame it as: “Can we try a different structure and see if it changes what we get?”
Before you start	Agree explicitly on the FERPA ground rules: what data travels, what stays in the room, what never enters a digital tool. Five minutes of that conversation protects your students and your team.

Has your team tried a structured data protocol? What made it work — or what got in the way? Drop it in the comments.

---

This is part of Accingo's Collaboration Hub — practical frameworks for the professional relationships that either make your job workable or make it worse.