Running Survey

Each learner writes two or three questions on a personalised topic (hobbies, travel, technology habits, opinions on a class issue), then surveys the rest of the class by mingling. They collect, tally, and report findings. The task rolls information-gap and opinion-gap into one: learners interview multiple partners, log data, and present a synthesis.

Procedure

1. Question design (7 min): In pairs, learners draft 2–3 survey questions on a chosen topic. Questions must include at least one Yes/No item and one open-response item (Why do you think that?). Teacher checks for grammatical accuracy before the mingle.
2. Survey grid: Each learner prepares a table with classmates' names down the left and their own questions across the top.
3. Mingle (10 min): Learners circulate, ask their questions, note answers. Goal: survey at least 8 classmates.
4. Analyse (5 min): Back at seats, learners tally numerical responses and cluster open responses into categories.
5. Report (varies): Depending on time, either a short oral summary to a small group, a one-paragraph written report, or a simple bar chart posted on the wall for a gallery walk.

Why it works

Unlike Find Someone Who, where the content is pre-specified, Running Survey gives learners authorship of the questions — so the language they generate afterward (I asked people whether they…, most of my classmates said…, surprisingly, only two people…) is genuinely theirs and genuinely contextualised. The task produces the report-verbs and quantifiers (most, half, a minority, roughly a third) that intermediate IELTS candidates specifically need for Task 1 writing, making this unusually transferable. Willis positions running surveys as a TBLT "listing-and-reporting" task sequence, with the data itself creating the need for the target language.

Variations

• Opinion survey: Questions all probe opinion (Do you agree that…? Why?). Report stage focuses on opinion-gap synthesis.
• Fact-finding survey: Questions are purely factual (How many hours of sleep do you get?). Report emphasises statistical language.
• Predict-then-survey: Learners predict class responses before surveying, then compare prediction to data — a strong source of genuine surprise.

Tips

• Vet questions before the mingle. Badly formed questions produce confused answers and a frustrating activity.
• Specify a minimum number of respondents (e.g., 8), which forces learners to keep moving rather than parking with one friend.
• Record useful reporting phrases on the board during the analyse stage. Delay error correction until after the report stage.