Week 10#
Communicating risk to visitors and new drivers#
Technically Accurate Is Not The Same As Actually Read#
A caution before we start#
This unit’s dataset doesn’t test who’s more likely to hit a kangaroo. We won’t claim visitors or new drivers are less careful — no source for that here.
What we can say: a driver with no local reference points needs a different brief. Regardless of who’s holding the wheel.
A fictional notice that fails#
“…wildlife hazards including but not limited to macropodidae may be encountered at any time… refer to clause 14(b)… Section 9 of the accompanying documentation pack…”
Read it at rental-counter speed. What’s the one thing you’re supposed to do?
Four stages, most warnings fail at #3#
- wrote it
- read it
- understood it
- acted on it
The fictional notice clears 1 and 2. It fails 3 — which makes 4 impossible.
A road sign has the same problem#
From Dr Fenn: a diamond warning sign clears “wrote it” and “read it” easily. It says nothing about when or what to do — so it can’t clear “understood it” for someone with no local calibration either.
Same failure, different medium.
The plain-English hierarchy#
Kangaroos and wallabies cross rural roads, especially at dawn, dusk and night. Slow down on rural roads during those hours. If you see one, brake steadily — don’t swerve. If you hit one, pull over safely and call [the after-hours number] — it’s on your key tag.
One sentence. Three actions. One pointer.
Rules for expressing risk#
- no invented numbers
- “conditions increase risk” ≠ “you will definitely hit one”
- no false guarantees either — “you’ll be fine” is equally unsupported
Three reader personas#
First rural drive — no mental reference at all.
Different road norms — nothing to unlearn, no local equivalent existed.
Jet-lagged and rushed — needs the fastest possible extraction of what matters.
Design personas, not a capability judgement.
Does comprehension prove anything about outcomes?#
Bagagiolo et al. (2019) tested migrant farmworkers reading agricultural safety pictorials in Italy — not road signs, not Australia.
68.8% / 35.4% / 32.4% comprehension across three groups. No pictorial hit the 85% ISO/ANSI threshold.
A structurally similar problem. Not evidence about kangaroo signage.
Next week#
Testing whether a brief is understood isn’t the same as testing whether it reduces collisions.
Week 11 applies the exact same caution to importing an entire intervention’s evidence from overseas.