Professional learningfor Coaches, Teachers, Librarians
Looking for the Checking
A looking-at-student-work protocol that asks one question of a class set: did students actually verify what the AI told them, and how can we tell?
All activities Professional learning
Student Voice Data Dive
Coaches and teachers design a short, privacy-safe student survey on devices, AI, and wellbeing, then read returned results with Notice-Wonder-Next and act on what students said.
Overview
Campuses make decisions about screens, AI, and phone rules every week, usually without asking the students who live with them. This session treats student voice as instructional data. Teams design a five-question anonymous survey and a short interview on when technology helps or hinders learning and wellbeing, check it against a privacy checklist, and then read a fictional set of returned results from a seventh-grade team using Notice-Wonder-Next. The protocol keeps the team describing before interpreting and reframing problems as design choices the team controls, and it ends with one change and a plan to tell students what their answers changed.
Objectives
- Participants will write an anonymous student survey of five questions or fewer, plus two interview prompts, that asks about learning and wellbeing without inviting sensitive personal disclosures.
- Participants will check the survey against a privacy-safe checklist and revise at least one question.
- Participants will use Notice-Wonder-Next to separate observations from interpretations in student survey and interview data.
- Participants will choose one instructional change the data supports and plan how to close the loop with students.
Materials
On paper
- Handout A: Privacy-Safe Student Voice Checklist (1 per team)
- Handout B: Returned Results, Mesquite Ridge Grade 7 (1 per person)
- Handout C: Notice-Wonder-Next Organizer (1 per team)
- Index cards for drafting survey questions; highlighters in two colors
On screen
- Optional: a form tool your district approves, set to not collect names or email addresses
- Optional: a spreadsheet or the form's summary view to chart results by class section
Before you start
- Read Handout B and complete Handout C yourself so you know which Notices are observations and which are interpretations in disguise.
- Find out your campus's rules for surveying students: whether parent notice is needed, which questions require approval, and what staff must do if a response raises a safety concern. Have that answer ready.
- If teams will build a live form, set the form tool so it does not record names, emails, or sign-ins, and test it once.
Step by step
-
10–5 min
Hook
Predict before you look
Read Question 3 from Handout B aloud without the results: "When I use AI for schoolwork, I check what it says." Each person writes a prediction of how seventh graders answered on a sticky note. Then reveal the results. Ask: "Where were you close? Where were you surprised? What else might we be guessing about our own students?"
Facilitator noteMany people predict either "they all use it and never check" or "they don't use it." The mixed real picture is the case for asking.
-
25–17 min
Explore
Design a five-question survey
Teams draft on index cards: five questions or fewer and two interview prompts for a small group of students. Give them a focus: "When does technology, including AI, help you learn in our class, and when does it get in the way of your learning or wellbeing?" Require at least one question about checking AI output, one about distraction or balance, and one open question. Each card should say what the team will do differently depending on the answer; a question with no possible action gets cut.
Facilitator noteSteer away from questions that invite disclosures a survey can't respond to, such as sleep problems, family situations, or mental health. Ask about school experiences the team can change: "When is the laptop in the way during class?" rather than "How much do you use screens at home?"
-
317–25 min
Practice
Privacy swap
Teams trade cards with another team and review them with Handout A. Reviewers mark every item Yes, Partly, or No and write one suggested rewrite. Teams get their cards back and revise at least one question. Ask: "Could any answer identify a student? Could any question put a student in an awkward spot?"
Facilitator noteCommon catches: a "what class period are you in?" question on a small team can identify students; an open question on phones at home can invite family details; a required open question pressures students who would rather not say.
-
425–45 min
Apply
Read the returned results with Notice-Wonder-Next
Say: "The Mesquite Ridge seventh-grade team ran a survey like yours. Here's what came back." Teams read Handout B and fill in Handout C. Notice (8 min): only statements someone else could verify on the page, highlighted in one color. Wonder (5 min): questions the data raises, in the other color. Next (7 min): one change the team controls, written as a design shift, not a student fix. A coach or team member facilitates and enforces the order: no Wonders during Notice, no fixes during Wonder.
Facilitator noteListen for Notices that are really interpretations ("students are addicted to their laptops"). Ask: "Where on the page does it say that?" Also listen for student-deficit framing ("they're lazy about checking AI") and reframe it: "What in our design makes checking hard or optional?" Strong teams notice that 41 of 71 students chose talking with a partner, that 33 were unsure of the class AI rule, and that one student has no home internet.
-
545–55 min
Create
One change and a "you said, we did"
Each team commits to one change supported by the data and writes three sentences on the back of Handout C: what students said, what we will change, and how we will tell them. For example: "Many of you weren't sure of our AI rule. We've posted it and we'll practice checking an AI answer together on Tuesday." Then they name the evidence that will show the change worked, such as a two-question follow-up survey or the next set of student work.
Facilitator noteClosing the loop is what makes the next survey worth answering. Students who see nothing change learn that their voice is decorative.
-
655–60 min
Debrief
What did students see that we didn't?
Ask the whole room: "What did the students' answers show that our observations of class alone would have missed?" Then ask: "Which response on Handout B would you want to follow up on, and how would you do it without breaking anonymity?"
Facilitator noteThe "no wifi at home" response is the classic case: the team can't find that student from an anonymous survey, but it can change the homework design for everyone, such as offline options and time in class.
Paper or screen
Unplugged
Run the survey on half-sheets of paper with no names, collected in a box or folder so no one sees who wrote what. Tally closed questions by hand with a partner, type or copy open responses onto a clean sheet so handwriting doesn't identify anyone, and run Notice-Wonder-Next on that sheet. Interviews happen as a short small-group conversation with notes taken by the coach, not recorded.
Digital
Build the survey in a district-approved form tool with name and email collection turned off, and share it through the LMS. The form's summary view makes it quick to compare sections or grade levels, and a coach can chart results for the team meeting. Remove any open response that could identify a student before sharing results beyond the team, and don't paste raw student responses into an AI tool to summarize them.
Does it need a screen? A form tool makes anonymous responses easier to collect and faster to chart across sections, which helps a team see patterns it would miss by hand. The interpretation, and the decision about what to change, stays a human conversation at the table.
Evidence of learning
What you should be able to see or collect if it worked.
- Draft surveys have five questions or fewer, and each question names an action the team would take depending on the answer.
- Revised surveys pass the Handout A checks for anonymity, voluntary answers, and no sensitive personal questions.
- Handout C shows Notices that can be verified on the page, kept separate from Wonders, and a Next written as a design change the team controls.
- Each team writes a "you said, we did" message and names the evidence that will show whether the change worked.
Adaptations
- Grades 3–5
- Use three questions with faces or thumbs for answers and one open question read aloud. Students respond on paper; the teacher reads the questions to the whole class.
- High school
- Invite two students to help write and pilot the survey. Their revisions usually make the questions clearer than the adults' drafts.
- Campus leaders
- Aggregate results across teams to inform a campus device or AI guideline, and share a one-page "you said, we did" with students and families.
Standards connections
Teams design a privacy-safe student survey on technology and wellbeing and interpret returned results with Notice-Wonder-Next.
Reflect
- How am I using data to inform and guide my coaching practices? (Coach ELE guiding question)
- When did I last make a decision about students' technology use without asking them?
- Which of my team's assumptions about students and AI did this data challenge?
Take it to your students
Run your revised survey with one class or grade level within two weeks, run Notice-Wonder-Next with your team, make one change, and tell students what changed. Repeat two questions a month later to see whether the change shows up in their answers.
Pairs well with
- TCEA Blog: Stop Fixing Students, Coaching for Equity Through Universal Design (Miguel Guhlin, 2026)The Notice-Wonder-Next protocol and the shift from student deficits to design barriers.
- TCEA Blog: Coaching Strategies, Part One (Miguel Guhlin, 2025)Student-centered, non-evaluative coaching that keeps the focus on results for students.
- Data Display Patterns field guideIdeas for charting survey results so a team can read them quickly.
