A new research poster from the School of Emergent Priorities links 238 shoppers' self-checkout trust surveys to their override-button logs over twelve weeks, finding no association between stated trust and observed override rate.
A 238-shopper field study finds override behaviour tracks trust tier not at all, across four store sites and twelve weeks
The School of Emergent Priorities has spent years asking what a stated attitude actually predicts once there’s a log to check it against. Its newest research poster picks a fittingly ordinary instrument: the trust survey handed to supermarket shoppers about the self-checkout’s item-recognition prompt.
Dr Iben Chikere, Dr Mirela Hanke, and Associate Professor Casimir Beng surveyed 238 shoppers across four store sites on how much they trusted the kiosk’s error prompt, then linked each respondent to twelve weeks of their own till-side override-button use. Override rate held inside a five-point band regardless of whether a shopper reported low, medium, or high trust in the prompt, and shopper frequency made no difference either way.
We built an instrument to measure something a shopper already does with their hand every week. It turns out the two were never talking to each other.
— Dr Iben Chikere, Lecturer, School of Emergent Priorities
“The trust score passed every check we ran on the survey itself,” said Associate Professor Casimir Beng, Lead of the Adaptive Metrics Lab. “It just doesn’t reach the till.”
The team is treating the null result as a finding in its own right rather than a study to redo: a shopper’s stated confidence and their override hand, it turns out, answer to different things entirely, and the poster reports the gap without proposing to close it.
The full poster is available from the University’s research repository under an open licence, doi:10.5555/slop.9pvjpl.