Direction-to-outcome mapping should match spatial intuition
Aliases: stick mapping · stimulus-response compatibility · population stereotype · CD compatibility
What it is
Push the stick forward: should the camera tilt up, or should the vehicle roll ahead? Nothing on the stick says, yet people arrive with an answer. Control-display compatibility of stick direction asks whether that default matches the system. It is about which way in the world a push should move, not about decomposing multi-axis components or whether deflection magnitude is analog or a switch—those can be tuned after the direction of the mapping is already right.
Why it happens
People drop body, device, and plant into one spatial frame, then fill gaps with a population stereotype: forward often reads as “increase in the direction I am looking” or “vehicle ahead”; up often reads as “rise.” Change the frame and the same motion inverts: a pan-head “push” is pitch or travel depending on whether the stick is being used as one’s own head or as a throttle. Under incompatibility the first move goes the wrong way. Where there is time to correct, the cost is a beat; where there is not, it is a reversal. Cursor mapping (up = cursor up) and vehicle mapping (up = nose up, so the picture actually sweeps down) can fight inside one product because both frames have been trained.
Studying it
Use stimulus–response compatibility tasks: same stick, compatible versus incompatible mapping, reaction time and first-move direction errors. A more ecological version gives no mapping rule, only a task (“put the lens on that tree,” “put the vehicle in that bay”), and watches the first push.
Independent variables: stick orientation relative to the body, plant as cursor versus vehicle, whether a verbal rule was trained, handedness. Dependent measures: whether the first direction is correct, number of corrections, reversal rate under an urgent command.
Practice can bring an incompatible mapping near compatible reaction times; errors return under urgency, distraction, or a hand switch. So besides a skilled block, measure the uninformed first trial and reversals under dual task.
Where it stops holding
Aviation and camera-follow trades have trained “odd” mappings into a professional default; rewriting their sticks to a lay stereotype scrambles experts. Handedness and cultural “up means on” also edit the stereotype; there is no single global table. When the plant has no stable “forward” (a wrapping menu, a continuously rotating head), spatial compatibility loses its anchor and labels or mode lamps have to carry the mapping. With several axes live, people may get the primary axis right and still run a second frame on the secondary; single-axis lab compatibility does not transfer automatically.
Applying it
- Write down the plant’s spatial frame (my hand, the vehicle’s nose, up on the screen) before deciding what the stick’s “forward / up” means; do not mix cursor and vehicle frames in one product without saying so.
- Follow that user group’s stereotype by default. If inversion is mandatory (a trade convention), put a non-removable direction mark on the boot and force a one-time direction check on first use.
- Verify: with no mapping briefing, give novices a task with a clear spatial goal and log the first push. If more than a small minority reverse, change the mapping or the frame explanation. Then probe experts under an urgent call or a secondary load, to see whether a trained inversion rebounds.