Skip to main content

6. Plan for mistakes

Decide: what do beginners get wrong in this task, and how will they find out?

You can't remove mistakes from learning, but you can predict most of them. Experts often no longer notice what makes a task hard, so ask before you build.

6.1 Find out the typical mistakes first​

Do

  • Ask an expert, or better, watch two or three beginners try the task. Ask: "What do beginners get wrong here? What do you check that they don't?"
  • Write down the two or three most common mistakes, and the step where each happens.

In ALF: the list decides the warnings in 6.2 and the items on your checklist. If you record a demonstration with Motion, mention each typical mistake in your voice explanation at the moment it could happen.

Avoid: relying only on the manual. It says what to do, not what goes wrong.

Why: experts skip steps and reasons they no longer notice. Asking about typical errors brings them back. Expert blind spot (Nathan & Koedinger).

6.2 Name the mistake at the step where it happens​

Do

  • Put one warning on the step itself, with the reason: "Don't … , because …"
  • Keep warnings for the real typical mistakes. Too many warnings are ignored.

In ALF: in the step's instruction, highlighted in a colour you only use for warnings: Write the instructions: Text.

The learner experiences: the warning comes exactly when it is needed, not on a page at the start that they've forgotten.

Avoid: a list of all warnings in the first step.

Why: feedback and guidance work best when they are specific and come at the moment of the task. Feedback research (Hattie & Timperley; Shute).

6.3 Show right and wrong side by side​

Do: where two things look alike but differ in one important detail, show both next to each other and point at the difference.

In ALF

Why: learners notice the important feature when they compare cases that differ in only that feature. Contrasting cases (Schwartz & Bransford), variation theory (Marton).

6.4 Give the reason, not just the rule​

Do: for every rule, especially safety rules, add one line on why. Learners who know why can handle the case the rule didn't mention.

In ALF: one sentence in the step's instruction: Write the instructions.

Avoid: "Always do X." without a reason.

Why: making the expert's reasoning visible turns rule-following into understanding. Cognitive Apprenticeship (Collins, Brown & Newman).

6.5 Use VR for the mistakes you can't afford for real​

Do: if a mistake is dangerous or costly in reality, show its consequence in VR, where it costs nothing. Never leave it to be discovered on the real object.

In ALF

Why: safe practice builds the confidence that keeps learners going. Mistakes are only productive where they are safe. Self-efficacy (Bandura), productive failure (Kapur).

6.6 Plan for wrong actions​

Do

  • A step only finishes on the right action, so a wrong action simply does nothing. Tell learners at the start: "If nothing happens, read the panel again."
  • Plan a short debrief after the session to talk about what went wrong.
  • If a supervisor is present, tell them which steps to watch: the ones from 6.1.

In ALF: steps that finish on the right action: Choose how a step finishes.

Why: feedback is most useful when it is specific and immediate, especially for beginners. Feedback research (Hattie & Timperley), Gagné's event 7 (provide feedback).

Not built in yet

ALF doesn't react to wrong actions yet: there is no message when a learner presses the wrong button. This is the biggest gap. Until it exists, the warnings in 6.2, a supervisor and the debrief do the job.