Part 1: Designing Training for Real-World Impact
Developed for the Failure Modes of Engineering (FeME) Project
Anelda van der Walt
Sheena O'Connell
| Welcome & warm-up |
| Exercise 1: System Map |
| Exercise 2: Theory of Change |
| Exercise 3: Learner Reality Check |
| Mid-session synthesis |
| Break (20 min) |
| How people actually learn |
| Exercise 4: Learning Outcomes |
| Exercise 5: Designing an Activity |
| Closing reflection |
In cross-team breakout rooms (3-4 people), introduce yourself briefly:
Speaking order: alphabetical by Zoom name
7 minutes
Think silently for 1 minute about your training.
You will have a chance to share one thing you feel excited about or one thing you feel uncertain about.
We're working from the outside in:
Everything today uses your own training.
Think of a training that didn't work — yours or one you attended.
What went wrong?
Was it the conten... or something around the content?
10 min · Own-team breakout
One team presents their system map highlights
You've mapped the terrain your training operates in.
Now: how does training actually lead to change?
A team ran a short training on analysing climate and environmental data for decision-making.
Participants worked through real examples. They could complete the exercises. The tools seemed useful.
There was energy in the room — people were already thinking about how they would use this in their work.
Then they went back. BUT... nothing changed.
They ran the same training at multiple sites — but none of them changed how they worked.
Why?
A Theory of Change helps you map how training is expected to lead to change — and where that chain can break.
Each link depends on an assumption.
These question marks are your assumptions.
If one fails, the chain can break there.
10 min · Own-team breakout
2 min · Individual thinking
Look at your chain. Find the point where it depends on something uncertain.
Write down one assumption that could break your training.
e.g. "Participants will have access to relevant data after the training"
Which assumption is most likely to break your training?
Vote — then we'll unpack your choices together.
"If you don't design for these assumptions, your training may work — but nothing will change."
You've mapped the system and the logic.
Now we zoom in on the people.
Everything you design will land — or not — based on how well it fits the humans who show up.
We were teaching basic web development to women online. The instructions said "open a text editor" — and many participants didn't know what that meant.
We hadn't met the learners up front. We'd aimed at one level of computer literacy. A different audience showed up.
We spent weeks creating content, but an unexpected audience showed up
Take 2 minutes of silent thinking before we go to breakout rooms.
10 min · Own-team breakout
Back at [TIME]
20 minutes
Three principles that should shape every design decision you make.
Working memory is smaller than you think.
Your job: reduce extraneous load so learners can focus on what matters. What are some strategies?
Memory is built when people pull knowledge out, not when you push it in.
People learn by talking, arguing, and building on each other's ideas.
Who has heard of learning styles?
Eg "visual learner", "auditory learner", etc
If you're an "auditory learner," could you learn to drive a car by listening to someone talk about it?
If you're a "visual learner," could you learn to play guitar by watching videos?
The evidence doesn't support matching instruction to styles.
What works: varying how you engage learners — because different activities build different kinds of understanding.
Let's focus on what learners should be able to do after the training.
Learning outcomes describe what learners will be able to do after the training, not what the training will cover.
This is your Theory of Change made concrete — where change becomes observable.
Don't overthink it. Let's use no more than 30 seconds.
| Topic description | Learning outcome |
|---|---|
| Understand climate data | Interpret a local dataset to identify two practical risks and explain them to a non-specialist audience |
| Learn how to use QGIS | Use QGIS to create a map showing flood-prone areas in a local region and explain the key risk patterns |
| Understand how data informs decisions | Use a dataset to compare two response options and justify a recommended action based on the evidence |
2 min · Individual thinking
Take your draft and improve it.
5 min · Own-team breakout
5 min · Whole group
Who has an outcome that shifted from vague to observable?
Read the original, then the revised version.
You've written outcomes that describe what learners will do.
Now design the activity where they practise doing it.
An activity is not a topic.
It's something learners do during the training that lets you see whether they can perform the outcome.
15 min · Own-team breakout
Pick one learning outcome from Exercise 4. Design an activity that helps learners achieve it. Or use one you already have
15 min · Whole group
If understanding were enough — this should work.
You don't need to finish it today.
The point was to experience what it feels like to design from outcomes — not from topics.
Take a moment to recognise what you've built so far.
The workbook will help you take this further, in your own time.
Who would like to share?
Find one resource that relates to your topic.
This could be:
Don't overthink this — one is enough. You don't need something perfect!
The goal today was not to master all of this — it was to see the full picture and find the parts that matter most for your work.
Teaching for Impact: Designing Effective & Open Training