Gallery Walk

A structure for feedback on teams' work. Every team lays its work out on its own table and picks a keeper, who stays with the work and the feedback form. The other team members go on tour past the other tables: at each one, first a minute of looking in silence, then feedback on the questions on the screen, while the keeper asks follow-up questions and writes it down. Every team visits all the other tables. Then everyone returns to their own table, the keeper reads out everything that was said about the work, and the team fills in the back of the form together: what we take on, what we deliberately do not, and how the teamwork went.

With the default settings the whole structure takes ± 30 minutes: 3 rounds, briefing and look-back included.

With the default times about half an hour for four teams. Two minutes to lay out the work and pick a keeper, five minutes per table, five minutes for the report and three for the team reflection. Because every team visits all the other tables, the number of teams sets the length: four teams are three visits and twenty-five minutes of steps in total, to which the preparation adds two minutes for arranging the room and two minutes for closing. Three teams: about twenty-five minutes; six teams: about forty. You set each step yourself, in steps of thirty seconds. The player only moves on when you click: a visit is over when the keeper has written down the last remark, not when the clock says so.

The keeper stays with their team's work for the whole tour. They do not defend and do not explain: they ask a follow-up question when a remark stays vague ("what exactly do you mean?") and write everything down, one line per visit under each question, including what they think is nonsense. Judging is for the team, later. After the last table they read everything out to their own team. It is not a role for whoever would rather not walk around: everything the team gets to hear about its work comes through the keeper's sheet.

Two to eight teams, and every team visits all the other tables: the number of rounds is the number of teams minus one. Four teams is the default and means three visits. With eight teams it becomes seven, and then it pays to shorten the visits. When everyone tours alone the number of rounds is the same, but each student chooses which piece of work they have not seen yet.

Leave the questions field empty and our three appear on the screen: what works well, what could be better, and what other ideas do you have. They work with any piece of work. It gets sharper with questions from the assignment itself: for an essay "is the thesis clear?" and "does the evidence convince?", for a design "can you tell what it is for without being told?". Two rules: every question must be answerable with an example from the work, not with yes or no, and three questions is enough. No time? Under ‘Set up & start’ the AI writes them for you: choose ‘AI writes them’, describe the assignment and year group, and you get three feedback questions to the same rules, which you read through before you start. Prefer Microsoft Copilot? Ask: ‘Write 3 feedback questions for students aged [age] who are looking at another team’s work: [describe the task, e.g. a poster about the water cycle]. The questions must be answerable in one sentence with an example from the work, not with yes or no. One question per line, no numbering.’

One sheet per team, printed double-sided. Side 1 is for the keeper, during the tour: a box per feedback question, one line per visit, and above each box a dotted line for anyone who put their own questions on the screen. Side 2 is for the team, after the report: two points you take on, what you deliberately do not take on and why, and one sentence about the teamwork: the same three things the reflection step on the screen asks for. The form is under Materials on this page, in three languages.

In a classic gallery walk the class walks past the work and looks, sometimes with sticky notes. What can go wrong there is that the feedback never arrives: the notes stay on the wall, the makers skim them, and nobody decides what happens next. This version adds three things: a keeper who catches the feedback and asks follow-up questions, a report in which the team hears all of it, and a reflection in which the team chooses what it takes on and what it does not.

In 1 goes, 3 stay teams trade results: each round one team member visits, collects what another team has and tells what their own team has, while the three who stay explain their work. The aim is that every team knows what the others found. In the Gallery Walk it is the other way round. One stays and the rest go: the work is not explained but looked at, and the aim is feedback the team can use. Rule of thumb: if the work needs enriching with what others know, take 1 goes, 3 stay; if the work needs to get better, take the Gallery Walk.

Three moments. At the start: help pick who keeps, and say out loud that the keeper does not defend but asks and notes. During the tour: walk along yourself, steer vague feedback towards the example ("which sentence?"), and note one pattern you hear at several tables: that is your whole-class minute after the reflection. At the report: hold on to the rule "hear everything first, react afterwards"; when a team starts refuting, ask "what exactly did they say?" That brings it back to the sheet.

Walking past each other's work is a classic classroom technique known as the gallery walk; Mark Francek described it in 2006 as a way to get discussion going about each other's work in science teaching. Spencer Kagan's cooperative learning has a related structure, Carousel Feedback, in which teams move from project to project and leave feedback; two of its principles are clearly visible here: simultaneity (all teams give feedback at the same time) and individual accountability (when everyone tours alone, every student phrases for themselves). The LudensFlow version adds the keeper who stays, the questions on the screen, the report and the choice on side 2. Why the keeper's sheet has to be the memory of the tour, we checked against Uncommon Sense Teaching by Barbara Oakley, Beth Rogowsky and Terrence Sejnowski.