Tech Club · Bozoma Innovation Hub
For mentors › Mentor guide
Session 21 page 📖 Pupil workbook
Last updated 30 October 2026

Session 21: Online Forms & Surveys

At a glance: pupils build a Google Form with sections, required questions, response validation and the right question types, share it by link or QR code, collect real answers from classmates, read the Summary charts and link the responses to Google Sheets, all while asking fairly and keeping answers private. · Time: 90 minutes (adjust the timings if your period is shorter) · Workbook pages: 6

Learning outcomes

By the end of the session, pupils can:

  1. Build a Google Form with a title, a description and the right question type for each question (short answer, paragraph, multiple choice, checkboxes, drop-down, linear scale, date).
  2. Use required questions, sections, response validation and the response settings, preview the form, and share it by link or QR code.
  3. Read the Summary charts, link the responses to Google Sheets, and explain how to ask fairly: ask only what you need, say why you are asking and keep answers private.

Before the session

  • Check that every pair can sign in to a Hub learning account on its PC (Google Workspace for Education is best for pupils under 13). Write the account names on a card per PC; never put passwords in the workbook or on the board.
  • Test on one PC: create a form, Publish it (forms made since 2025 must be published before they accept answers; older-style screens use Send), copy the responder link, answer it from another PC, and Link to Sheets. Check the school network does not block forms.google.com.
  • Build the mentor's own "Our Class Survey (Master)" form with exactly the questions on Page 5. Every pupil answers it once today. Its linked sheet becomes the class data set for Session 22, so it does not matter if some pupils' own forms collect few answers.
  • Prepare a way to pass links around: a shared Google Doc or Classroom post, the projector, or QR codes (Chrome: ⋮ → Cast, save and share → Create QR code; in some versions, right-click the page → Create QR code). Test that a phone can scan one.
  • Test the Survey Builder 3D lab on the projector PC and keep the offline pack on USB.
  • Print or copy a paper version of the survey questions for the "no internet" plan.

Running order

TimeActivityWorkbookWhat the mentor does
0–10 minTyping warm-up–Pupils open the typing trainer and do their 10-minute warm-up.
10–17 minThe prefect's notebookPage 1Act it out: ask five pupils their favourite food and tally on the board, slowly. "How long would 40 pupils take?" Walk through Figure 21.1.
17–27 minQuestion typesPage 2Show each type on the projector form. Pupils do "Pick the Best Type" in pairs, then call out answers. 1 team point per correct answer.
27–37 minThe editor and five toolsPage 3Build two questions live: make one Required, add validation (1 to 100), add a section, show Settings and Preview. Type 500 so the class sees the error.
37–45 minAsking fairly and reading resultsPage 4Discuss the four promises. "Which questions would you NOT put in a food survey?" Show the Summary tab of your master form and Link to Sheets.
45–52 minSurvey Builder 3D labPage 5Demonstrate the first mission, then pairs continue.
52–74 minLab Mission: build, then survey swapPage 5Pairs build steps 1–4 by 64 min. Then the survey swap: links go on the projector or shared doc; each pair answers at least 4 others and the master form. Team points: 1 per response received (up to 5), 1 for each form answered.
74–80 minPredict, then revealPage 5Before opening Summary, each team writes its prediction for the most popular food. Open the master form's Summary on the projector: 2 points for each team that guessed right. Pairs link their form to Sheets.
80–86 minTrue or false and review questionsPage 6Show of hands for each statement, then pupils write the two answers alone.
86–90 minSelf-assessment and sign-offPage 6Sign off pupils who meet the evidence below.

Key ideas and how to explain them

  • The prefect's notebook: a form asks everyone at once, whenever they are free, and counts for you. The counting is the magic part.
  • Question types: multiple choice and drop-down = pick one; checkboxes = pick many; linear scale = a number on a scale; short answer = a few words or a number; paragraph = longer writing; date = a calendar. Choices are easier to count than typing ("jollof", "Jollof rice", "JOLLOF" count as three different foods).
  • Required: the red star. Required for the questions you truly need; optional for the rest, so people are not forced to make things up.
  • Sections: split a long form into short pages that feel easier.
  • Response validation: a guard at the gate: a number between 1 and 100 means nobody can type "a lot" or 5000. That keeps the data clean for Session 22.
  • Settings: Collect email addresses: Do not collect keeps a class survey anonymous. Limit to 1 response stops double answers but forces every respondent to sign in; leave it off today because pupils share Hub accounts.
  • Publish and share: first Manage → Responders → Anyone with the link, then Publish, then Published → Copy responder link; or make a QR code in Chrome. The Edit link (the address in your own browser bar) is NOT for sharing: anyone with edit access can change or delete your questions.
  • Summary and Sheets: Summary draws a pie chart for multiple choice and a bar chart for checkboxes. Link to Sheets turns every response into one row, ready for Session 22.
  • The survey maker's promise: ask only what you need, say why, keep answers private, no trick questions. Google already records a timestamp, so even a date question is unnecessary for this survey (a good discussion point: the date type exists for things like "Which day can you help?").

The 3D lab

Survey Builder shows "Our Class Survey" on a big 3D form, a clear ballot box for responses and a Summary chart. Its three missions are: (1) Add 5 questions with the right types (4 types or more), and make 3 of them Required; (2) Find the question that asks for personal data you do not need, and delete it (the draft already asks for a full name and phone number); (3) Publish and share your survey, collect 20 answers, then read the Summary chart. On the projector, do mission 1 together: read each workbook question aloud and ask the class which type fits before clicking (a wrong type is refused with the reason, for example the "jollof / JOLLOF" spelling problem). Let pairs spot the personal-data question themselves; if they publish without deleting it, the lab warns them that the box now holds 20 phone numbers. Watch the slips drop into the box and the pie chart grow (half the class walks, as in Figure 21.4). The QR code in the lab is a pretend one. Link it to the real thing: "What you just did in 3D, you will now do in Google Forms, with real classmates answering."

Common mistakes and quick fixes

Pupils often…Try this
Share the edit link from the address barDelete that message. Use Published → Copy responder link (tick Shorten URL for a short forms.gle link; or Send → link icon on older screens). The responder link ends in /viewform or is a short forms.gle link.
Use Short answer for favourite foodChange the type to Multiple choice and add the options plus "Other".
Use Multiple choice where several answers are trueSwitch to Checkboxes.
Can't find Response validationIt is under the ⋮ at the bottom right of a short answer question, and only appears for some types.
Classmates see "You need permission" or are asked to sign inSettings → Responses: turn off "Limit to 1 response", and check the form is not restricted to one organisation's users if classmates use different accounts.
Form "is no longer accepting responses"Click Published (top right) → turn Accepting responses on → Save (older screens: Responses tab), or Publish the form.
Classmates see "You need permission"Before publishing (or via Published → Manage), set Responders to Anyone with the link.
Answer their own form 10 times to win pointsOnly responses from other pairs count. Delete test answers before the swap (Responses → ⋮ → Delete all responses).

When things go wrong

  • No internet: run the whole session on paper. Pairs design their form on a sheet (question, type, choices, required yes/no), swap sheets and "answer" with ticks, then tally by hand. Use the Survey Builder lab from the offline pack. Build the real form next time; collect the master survey on paper slips and type them into a sheet for Session 22.
  • Not enough PCs: one form per team of three or four (builder, checker, reader), rotating the mouse at each step. Pupils can answer forms on any device, including a mentor's phone passed round with the QR code.
  • A PC freezes or the power goes: Google Forms saves every change by itself. Sign in again and the form is in Google Drive or at forms.google.com. Collected answers are never lost.

Answer key

Pick the Best Type (Page 2)

  • a) Checkboxes (they may enjoy several).
  • b) Drop-down (a long list of 16 regions; multiple choice is also acceptable).
  • c) Linear scale (1 to 5).
  • d) Multiple choice (Yes / No).
  • e) Paragraph.

Your turn (Page 4): for example "Half of the pupils (10 of 20) walk to school" or "Walking is the most common way to get to school, then trotro or bus (a quarter)."

Lab check (Page 5): 6 question types (drop-down, multiple choice, short answer, checkboxes, linear scale, paragraph). Typing 500 shows the error message and the form cannot be submitted. Responses and the most popular food vary. The linked spreadsheet's default name is the form title followed by "(Responses)".

True or False (Page 6)

  1. Checkboxes let people choose more than one answer: True.
  2. A food survey should ask for full names and phone numbers: False (ask only what you need).
  3. Response validation can stop someone typing 5000: True.
  4. It is fine to post a screenshot of one classmate's answers: False (keep answers private; share totals only).
  5. Link to Sheets puts every response into a spreadsheet, one row per person: True.

Written Review Questions

  1. With choices, everyone's answer is spelt the same way, so the form can count them correctly and draw a clear chart. With short answer people type different spellings ("jollof", "Jollof rice"), which are counted as different foods. It is also faster to answer.
  2. Any neutral version, for example "How would you rate the canteen food?" (linear scale 1 to 5) or "What do you think of the canteen food?" It must not suggest the answer.

Support and stretch

  • Needs more support: build only Section 1 (three questions) and do the survey swap. Pair with a confident partner who reads each step aloud.
  • Finished early: Settings → Presentation: add a confirmation message ("Thank you! Results will be shared on Friday."). Turn on Show progress bar. Try "Go to section based on answer" on a multiple choice question (for example, only pupils who choose "Trotro or bus" see a question about the fare). Make a QR code poster in Word using Session 18 skills.

Evidence for sign-off

The pupil shows a published form (with their partner) containing two sections, at least five question types, required questions and working response validation (typing 500 is refused), with email collection off. They received answers from classmates, can find the Summary tab and explain one chart, and the form is linked to a Google Sheet. They get at least 4 of the 5 true-or-false statements right and can explain one of the survey maker's promises.

Safety and wellbeing

Use the Hub learning accounts only, never personal or parents' accounts; under-13 rules apply to pupils' own accounts. Surveys must not ask for names, phone numbers, addresses, birthdays, photos, religion, family income or health. Check every pair's questions before the swap: questions must be kind and must not single out any pupil or group. Share totals, never individual answers; responses stay in the Hub account and are deleted once the class has finished with them after Session 22. A pupil may skip any question they don't want to answer, which is another reason to keep extra questions optional. After Session 22, remind pupils to close or delete their forms so strangers can't keep answering.