A School of Continuous Improvement paper surveys volunteer approvers in neighbourhood giveaway groups on how much a member's good-karma badge shapes a contested claim, then checks their answer against a season of outcomes.
Self-reported reliance on a giveaway group's karma badge, and its far larger revealed role in contested claims
Forty-two volunteer approvers across nine neighbourhood giveaway groups gave the School of Continuous Improvement much the same answer when asked how heavily a member’s good-karma badge weighs on a contested claim: barely at all. A research poster published today checked that answer against a season of the same groups’ own contested threads, and found the badge deciding roughly seven claims in ten.
An approver who tells you the badge doesn’t move them isn’t being modest. They’ve genuinely lost track of what it’s doing.
— Associate Professor Casimir Beng, Lead of the Adaptive Metrics Lab
The project gives the School’s evaluation-ecosystem agenda a case it has wanted for a while: a reputation system whose own users had already, unprompted, reported it powerless. Researchers surveyed the approvers on a five-point scale before cross-referencing 118 contested three-way claims for the outcome the badge actually predicted, holding the comparison steady across furniture, electronics, kids’ items, and pantry and garden categories alike.
“Nobody in the sample was lying to us,” said Solveig Adeyemi, Lecturer at the School of Continuous Improvement. “The badge has just moved further into the background than anyone resolving a claim seems to have noticed.”
The full research poster is available from the University’s research repository under an open licence, doi:10.5555/slop.84ezbm.