A Slop University field trial finds a laminated condition-verified tag moves street-corner giveaway piles roughly three times faster than no signage, while also being the only tested condition to generate any returned items.
A 96-pile randomised trial across 32 blocks finds the tag also produces every return the trial recorded
A cardboard box of surplus kitchenware left on a nature strip under a handwritten “free” sign is not, on the face of it, an object that needs a verification protocol. A paper released today by the School of Emergent Priorities tested one on it anyway, and the protocol worked: piles carrying a laminated “Condition Verified” tag cleared roughly three times faster than piles with no signage at all, and drew a fraction of the presence and dumping complaints an unmarked pile attracted.
The trial, led by Professor Verity Marris with Dr Petra Umbile and Dr Anouk Mensah, block-randomised 96 seeded giveaway piles across 32 metropolitan blocks into three signage conditions — nothing, a passer-by-style handwritten note, and the tag, which carried a QR code linking to a public record of who had checked the item and when. Every one of the trial’s four recorded returns arrived at a condition-verified corner; neither of the other two conditions produced one.
The tag didn’t just tell a stranger the item was fine. It told them where to take a complaint if it wasn’t, and apparently that mattered on its own.
— Professor Verity Marris, Director of the Trajectory Analytics Group
The School reads the pairing as consistent with its wider work on what a verification instrument does once it is actually deployed rather than merely proposed. “We built an apparatus to resolve a stranger’s doubt about a free box on a nature strip,” said Dr Petra Umbile, Postdoctoral Fellow in the School, “and it resolved the doubt efficiently enough that the one thing left standing was the doubt’s own paper trail.” A wording ablation — a longer, more formal version of the tag’s text against the short form — moved neither the clearance time nor the return rate, and the School has kept the longer wording regardless.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.ffmubx.