Another Look at Gradual Release: What Does “I Do” Really Mean?

by Michelle Yzquierdo

A seventh-grade teacher is solving a two-step equation at the board. “What do we do first?” 
A few students shout out, “Subtract four!” 
The teacher writes it down. “Now what?” 
Another student shouts out: “Divide by two!” 
The teacher finishes the problem and moves to the next one.

This looks like direct instruction. The teacher modeled a problem at the board. But look closer at who did the thinking. It was not the teacher, nor was it most of the class. Rather, three or four students carried the cognitive work for the whole class, and everyone else watched. 

This is a math example, but the subject doesn’t matter. The same thing happens when an ELA teacher annotates a paragraph while students call out the theme, or when a social studies teacher pulls up a primary source document and asks, “What do you notice?” I see this pattern frequently when I visit classrooms, across all grade levels and all content areas. It seems to be the norm, and not an outlier.

This is the most common way the “I Do” phase of gradual release breaks down. When P. David Pearson and Margaret Gallagher first defined gradual release of responsibility in their original 1983 research, the modeling phase (“I Do”) was built on one simple premise: the teacher holds all of the responsibility for the work, and the student holds none of it yet (Pearson & Gallagher, 1983; full text via ERIC). “I Do” works when the teacher carries 100% of the thinking out loud, without student input. The moment a student supplies a step, the model stops being a model.

Why This Matters More Than It Seems

On the surface, student input during “I Do” looks like a small technicality. Does it really matter if a few kids call out answers while the teacher writes? It does matter because of what students are supposed to walk away with from “I Do”: a complete, internal mental picture of what expert thinking sounds like, start to finish, including the deliberate checking and self-monitoring that strong thinkers build in along the way.

When a teacher solves the equation alone, students should hear things like: “I want to get the variable by itself, so I’m going to undo whatever was done to it last. Wait, I need to check the order of operations here. What happened to x first?” 

That narration is the whole point. It shows students how to decide the next step, not just which step comes next. A student who only hears “subtract four” from the kid in the front row learns a procedure for this one problem. A student who hears the reasoning behind it learns how to approach the next problem, and the one after that, and the one after that. 

The same is true whether it’s a math step, a student calling out, “this document shows bias,” in U.S. history before the rest of the class has had a chance to look closely at the document themselves, or another who shouts out “personification” before anyone else has the chance to look at the example of figurative language on the screen. 

Once a student calls out a step or an answer instead of the teacher doing the think-aloud, the reasoning disappears. The teacher writes down a correct answer, but the thinking that produced it lives in the head of whichever student said it out loud, not in the thirty other heads in the room. Multiply that across a school year, and you get a class full of students who’ve watched hundreds of problems get solved but still have trouble when they are handed new problems to solve alone. They were never actually shown how the decisions got made. They were shown, by other students, what the decisions were.

What a True “I Do” Actually Sounds Like

Let’s go back to that same equation, modeled the way “I Do” is meant to work:

“I have 2x + 4 = 14. My goal is to get x alone. Right now, there’s a plus four attached to the x variable, so I’m going to undo that. First, I’ll subtract four from both sides. That gives me 2x = 10. Now x is being multiplied by two, so I’ll undo that by dividing both sides by two. X = five. Let me check that. Two times five is ten, plus four is fourteen. That’s the same as what I started with, so I’m confident in this answer.”

No pauses for input. No “What’s next?” Every decision, what to undo first, why, and how to check it, comes from the teacher, out loud, the whole way through. Students aren’t supplying ideas or answers during the model. They’re listening to the entire cognitive process, from beginning to end.

This is harder than it sounds, especially when call-outs and shouter-outers make a room feel more engaged than it actually is. A few voices create the impression of overall engagement, but it’s really only the same three or four kids carrying the room. When you revamp “I Do” to look and sound like it is supposed to, it can feel like you’re “losing” the energy when you stop allowing students to shout out or stop calling on them. However, that energy has a home in the next step of the lesson, not inside the “I Do” model itself.

So what are students doing during “I Do” if they’re not calling out anything? They’re tracking the problem on their own paper as the teacher solves it out loud, capturing the reasoning from the teacher in addition to the answer. This gives them something to draw on and refer to when they get to the next step of gradual release. And quiet students do not mean passive students. Quiet means the thinking is happening in everybody’s head, and not just in the heads of the three or four shouter-outers.

A Small Addition Worth Considering

Once your “I Do” is solid, there’s a second idea worth considering: the jump from guided practice (“We Do”) straight to independent practice (“You Do”) is often too big for some students who need more support before they are ready to fly solo. Consider expanding “We Do” into two substeps: a partner “We Do” and a solo “We Do”. Partner work gives students the chance to catch mistakes and an audience for their own think-aloud before they are on their own. 

The revamped gradual release process looks like this:

• I Do – 100% of the cognitive load sits with the teacher. The teacher thinks aloud through the entire process, start to finish, with no student input, making expert thinking visible.
• We Do (partner, supported) – Students work the task together in pairs, talking through their thinking the way the teacher just modeled it. The teacher circulates through the room, listening in on conversations, asking and answering questions that surface, and noting which pairs are solid and which are struggling. This is the first real data point on who’s ready to move on.
• We Do (independent, supported) – Students work the task on their own, but the teacher is still actively circulating, checking in, asking clarifying questions, and redirecting students who’ve gone off track. Nobody is left to struggle silently through an entire problem before getting help. This is where the teacher figures out, student by student, who’s actually ready for true independence.
• You Do – True independent practice. No partner, no teacher prompting, no checking in. This is the phase that tells you what a student can truly do alone. Based on the data the teacher collected during the “We Do” phases, this is also the perfect opportunity for a teacher to pull a small group for a quick reteach.

Try It This Fall

When you’re back in the classroom in the fall, resist the urge to solicit student responses during the “I Do” phase of your lesson. Let the silence sit while students process what you are doing and saying. Narrate every decision, including the parts where you double-check yourself. You might find your room a little quieter than you’re used to during this part of the lesson, and that’s OK. Students will have plenty of opportunities to do their own thinking, processing, and talking during the partner and the solo “We Do” components.