From the Gap to the Mechanism

Using a torn paper model, time-lapse evidence, and student predictions to turn 'the wound got better' into a question about what cells are actually doing.

In this chapter, you will see...

What stays protected

Return to the injured foot • visible evidence of healing • a shrinking gap • cell-level prediction • multiple competing mechanisms • modeling before explanation • navigation into cell division

What is still being tested

How much of the DQB revisit is worth keeping; whether the time-lapse needs a full observation sheet; how much modeling should happen before students see candidate mechanisms; and whether removing some scaffolds gives students better ownership without sacrificing the quality of the reasoning.

Lesson 8: From the Gap to the Mechanism

Why this is one of the OpenSciEd lessons I mostly keep—and why my best modifications make the invisible problem more concrete before asking students to explain it.

By William Pirl | Unit 6.6: Injuries & Healing | PIRL Experimental

Lesson 8 is different from several of the lessons that come before it because I do not feel a strong need to replace the science story.

The published lesson already makes a useful conceptual move. Students return to the original injury, watch skin heal over time, and try to explain how a visible gap in tissue becomes whole again. The lesson deliberately stops before giving them the mechanism. They should leave with a stronger model of what healing must accomplish and a new problem: where do the new cells come from?

That basic progression works for me.

The lesson is strongest when the gap becomes something students can actually think with.

Where my modifications begin

My changes are less about changing the destination and more about making the middle of the lesson more concrete, faster, and more generative. I want students to see the gap, manipulate the gap, predict what could fill it, and argue about competing explanations before Lesson 9 introduces the mechanism more directly.

What OpenSciEd is trying to build

OpenSciEd frames Lesson 8 as a one-day problematizing and investigation lesson. Students revisit the healing timeline and DQB, observe a time-lapse of a skin wound healing, model what they think happens at the microscopic scale, revise their definition of healing, and generate new questions that lead into cell growth and splitting.

The science target is appropriately narrow. Students are not yet supposed to explain cell division. The lesson is designed to make that mechanism necessary. They can observe that the wound becomes smaller, that skin eventually covers the damaged area, and that the healed tissue must contain cells. The missing explanation is how that happened.

That is exactly the kind of uncertainty I want to preserve.

Visible healing is evidence. The cellular mechanism is still a hypothesis.

Modification #1: Make the gap physical

The most useful modification I have developed is also one of the simplest.

I give students a layered cross-section of the tissues involved in the injury. They rip the image in half to represent the damage and tape the two halves into their notebook so they can open them like barn doors.

Figure 2. The layered tissue image used as a physical model of the injury. Students rip the image through the tissue layers and tape the two sides into the notebook so the missing space can open like a pair of doors.

The missing space between the two halves becomes the problem.

From there, I can keep the task very open. Students can See-Think-Wonder, jot bullet points, make educated guesses, or draw intermediate stages. The exact product matters less than the question they are now forced to confront: how can all of these layers become continuous again?

Healing stops being 'getting better' and becomes 'filling a gap.'

One surprisingly useful role ChatGPT has played in this project is helping me turn rough, hand-drawn classroom ideas into cleaner instructional visuals. I can sketch something quickly in a notebook—sometimes just circles, arrows, labels, and a few colors—and upload a photo of it to ChatGPT. From there, ChatGPT can interpret the structure of the drawing and help recreate it as a more polished digital image while preserving the original teaching idea. That is useful because many of my best classroom models begin as something I draw spontaneously while explaining a concept. I do not need to be a graphic designer, and I do not need to abandon a useful visual simply because the original version looks like something drawn by a teacher with a marker five minutes before class.

I have found this especially valuable for science diagrams. ChatGPT can help clean up tissue models, simplify complicated anatomy, isolate individual systems, remove distracting details, and create consistent versions of images I want students to use repeatedly. I can also iterate conversationally: make the skin layer thinner, simplify the cells, remove labels, add a missing tissue, make the model less realistic, or match the style of another diagram. The important part is that I am still making the instructional decisions; ChatGPT is helping me translate those decisions into a usable visual product much faster than I could on my own. For teachers who have good ideas but limited time, artistic skill, or access to professional design tools, that ability to move from hand sketch → conversation → polished teaching visuals can be remarkably powerful.

Modification #2: Work backward from healed skin

Students are usually able to reason backward from the intact endpoint more successfully than I expected.

The healed model gives students a concrete endpoint. If there is continuous skin tissue at the end, and there was an open gap earlier, then something must have happened in between. This shifts the question from 'What does healed skin look like?' to 'What sequence of changes could have produced this final state?'

That backward reasoning is valuable because it lets students use what they already know about tissue organization without pretending they already know the mechanism.

Modification #3: Model the bleeding and scab stages with the systems already learned

The open wound and scabbed wound give students an opportunity to bring several earlier lessons back into the healing story.

Figure 3. Student notebook models of a bleeding wound and a scabbed wound. The model integrates the gap in skin with blood, platelets, plasma, red and white blood cells, and the developing scab.

This is one of the strengths of putting Lesson 8 after the blood, nerves, and tissue work—or, if the sequence is adjusted, at least after students have enough background to recognize those structures. The wound is not simply missing skin. It is a disrupted system involving tissue, blood vessels, blood components, and sensation.

Students can therefore build a model that is richer than the generic 'cells move into the gap' story, while still staying inside the evidence they have accumulated.

The more systems students can see in the wound, the less healing looks like a single-event process.

Modification #4: Ask for the missing middle

The most important model may not be the bleeding wound or the healed skin. It is the stage between the scab and the intact tissue.

That is where I ask students to predict what healing looks like at the cellular level.

They might propose that cells at the edges grow larger. They might think nearby skin cells move into the gap. They might think more skin cells are produced. They might suggest that other kinds of cells transform into skin cells. At this point, I do not need to settle the disagreement.

Modification #5: Vote, debate, and propose alternatives

One of my hallmark time-saving moves in OpenSciEd is to stop requiring every student to independently invent every possible model. When OpenSciEd provides or anticipates several common ideas, I can put those possibilities in front of students and make the task evaluative rather than purely generative. I tell them directly: these are some of the ideas OpenSciEd expects students to propose. Which one do you think is best? Which one is weak? What evidence would you need? Do you disagree with all of them? Can you create another option?

This usually produces good debate while saving the substantial amount of class time that can disappear into drawing. It also lowers the barrier for students who have a reasonable scientific idea but struggle to represent it clearly on paper.

Sometimes giving students the models creates more thinking than asking them to draw the models.

Figure 4. “How Does Skin Fill the Gap?” OpenSciEd’s model presents several competing hypotheses for what cells might be doing as damaged skin heals, including cells growing larger, nearby skin cells helping form new skin cells, or other cells contributing to the repair. The purpose is not to give students the correct answer, but to make alternative mechanisms visible so they can compare, debate, and later test them against new evidence. Source: OpenSciEd Unit 6.6, Lesson 8 Teacher Edition.

This year's experiment: remove some of the scaffolds I added

The irony is that the modifications that made this lesson reliable may now be the pieces I need to partially remove.

Between the time-lapse, See-Think-Wonder, ripped-layer model, prediction, candidate mechanisms, voting, debate, and proposed alternatives, I have enough material for a strong 78-minute lesson. Under the new schedule, that is too much—and I am not convinced every scaffold is still earning its time.

What I want to test

The next rebuild may require me to trust students with more of the uncertainty.

OpenSciEd vs. my rebuilt Lesson 8

What I am protecting, compressing, and still questioning

The version I want to teach this year

A possible 66-minute Lesson 8

The bigger point

Lesson 8 works because it does not try to teach cell division before students need cell division.

The visible wound gives them evidence. The ripped paper gives them a concrete gap. The models give them competing mechanisms. The healed skin gives them the endpoint. The only thing missing is an explanation for how more cells appear.

The best setup for a new mechanism is a problem the old model cannot solve.

Sources & design notes

Primary curriculum source: OpenSciEd, Unit 6.6, Lesson 8 Teacher Edition, “What happened as the skin on top of the foot healed?” Version dated 2/9/22. The published lesson uses a skin-healing time-lapse, an initial cell-level model, a revised definition of healing, and navigation toward figuring out how new cells form.

Teacher-created modifications shown here include the ripped layered-tissue notebook model, bleeding/scab/healed student modeling, candidate mechanism voting and debate, and the planned reduction of scaffolds to protect more student modeling time.

LESSON 8 | CONDENSED MODIFICATION GUIDE

Unit 6.6 • Injuries & Healing

PROTECT: gap • time-lapse evidence • cell-level prediction • competing mechanisms • model revision • navigation to new-cell production

RECOMMENDED TIMING

Adapted from OpenSciEd Unit 6.6, Lesson 8 • Teacher-use summary

Lesson 8: Practical Modification Guide

Skin Healing, Gaps & Predictive Models

What this lesson really needs to accomplish

• Students reconnect to the original injury and the idea of damaged tissue.
• Students observe a clear sequence of visible healing.
• Students represent the wound as a physical gap.
• Students include skin cells and relevant blood components in their model.
• Students propose one or more mechanisms for how the gap becomes filled.
• Students compare and critique competing mechanisms.
• Students leave with a question that naturally sets up cell division.

What I have done in the past

1. Ripped layered-tissue model

Students rip a body-layer image in half and tape the halves into the notebook like doors. The missing space becomes a manipulable representation of the injury.

2. Time-lapse + See-Think-Wonder

Students watch visible healing over time and connect what they can see to what might be happening at the microscopic scale.

3. Bleeding → scabbed → healed models

Students work backward from intact skin and model the wound at several stages, integrating skin, blood vessels, blood cells, platelets, plasma, and nerves as appropriate.

4. Candidate mechanism vote

Students evaluate possible explanations for how the gap fills rather than spending the entire period generating polished drawings from scratch.

What I am considering this year

1. Less DQB processing

Use one or two questions to reconnect to the unresolved healing problem, then move directly into the gap.

2. Less observation scaffolding

Keep the video, but reduce the amount of written noticing so students have more time for the cell-level model.

3. More modeling before options

Let students create the missing-middle mechanism independently before showing candidate models.

4. Use candidate models for revision

After students commit to an idea, use the OpenSciEd-style possibilities to challenge, defend, or revise the original model.

5. Protect disagreement

Do not collapse the class to one answer. The unresolved disagreement is what makes Lesson 9 necessary.

My current recommendation

Keep the rip-and-flap tissue model, the healing time-lapse, and the student missing-middle model. Compress the routines around them. This year, shift more time toward independent prediction before exposing students to candidate explanations.

What to cut first

• Full Scientists Circle DQB processing.
• Moving answered questions to new posters.
• Extensive noticings at every time point.
• Multiple new class posters.
• Two-round talking-stick protocols.
• Lengthy 'I used to think / now I think' writing.
• Formal renaming of posters and driving questions.
• Extended generation of new DQB questions.

What not to cut

• The physical gap.
• The time-lapse or other visible evidence of healing.
• A cell-level prediction.
• At least two competing mechanisms.
• Student revision or defense of a model.
• The idea that healed tissue contains cells where a gap once existed.
• The unanswered question of how new cells are produced.

One rule for a teacher new to the curriculum

Do not teach cell division until students have a reason to need cell division.