Skip to content

Module 03 / Lesson 03

Separate Observations From Assumptions

Write down what happened, name what you’re guessing, and pick a small way to check.

SanX InstructorIntroduction
Read the transcript

Introduction

Four people stopped on a workshop booking form. Let’s work out what we know before deciding what to change.

  1. Separate the event from the explanation

    You watch eight people use a workshop sign-up form. Four pause at the step asking them to confirm their email address. You can record the pause because you saw it. You can report the number because you counted it. Someone on the team says, “They don’t trust the site.” That might be true. They may also have missed the email, typed the wrong address, or expected the page to refresh. We need another check to say why they stopped. Write each claim on a separate line. Put the count beside the event, and keep your explanation under assumptions. That makes it clear which part needs more evidence.

  2. Check your best guess

    Watch what happens next. Ask someone to think aloud after they submit the form. One person says, “I don’t know if I’m booked yet.” That gives us a clearer idea of what to check. It doesn’t tell us why every other person paused. Try a clear message that says, “Check your email to finish booking.” Keep the rest of the page the same so you can pay attention to the message. Watch whether the next people know which step to take. Use the same task for the next group. You want to know whether people can finish the booking, so ask them to get a confirmation without asking you what to do.

  3. Update the notes

    If people get past the new message faster, note the result and how you measured it. If they still get stuck, watch which step causes the trouble. That tells you what to investigate next. When you hand this to a teammate, include the original observation, the wording you tried, and what you still need to check. That gives them an evidence trail they can follow without guessing what happened.

Recap

Write the observation first, label your guess, then pick a check that could prove it wrong. Try naming the two in the quick check.

Listen to the audio only

Try it yourself

Which line can you record as an observation from this session?

Label observations and assumptions separately, then test an explanation with a small check.

Next, use that habit to check something an AI assistant tells you.

Next lesson