A new Slop University paper attempts to reproduce, on ordinary community apps, a capture effect first observed in a strip-mall till-variance index, finding no anticipatory change in reply speed across 58 platforms, and reports the School's decision to roll the index out anyway.
A Responsiveness Index announced across a group chat, a forum, a booking app and a review-score marketplace produces no anticipatory change in reply speed, and the School adopts it regardless.
An index that changes nothing before it starts is still worth building for the day it changes something.
— Dr Renke Sabel, Senior Lecturer and Convenor of the Indicator Commons
A paper from the School of Continuous Improvement reports what happened when Dr Bram Ntuli, Associate Professor Casimir Beng and Dr Renke Sabel tried to reproduce, on four ordinary community apps, an effect the University first noticed in a strip-mall till. Announcing that a new Responsiveness Index would soon decide who counts as a trusted neighbour changed nothing about how quickly volunteers actually replied to a dispute — and the School has decided to introduce the index regardless.
The team built the Responsiveness Index to travel: the same formula, drawn from how fast a dispute gets a reply and how well that reply is later rated, applied without adjustment to a street group chat, a neighbourhood forum, a booking app, and a review-score marketplace across one region. Fifty-eight platforms took part, thirty-one of them told in advance that the index would soon govern eligibility for a portable community-trust badge; reply speed on those platforms moved no differently from the twenty-seven that were told nothing at all.
“A null result that holds across four very different platforms is still a result,” said Dr Ntuli, who led the study. The School regards the finding as exactly the kind its evaluation-ecosystem agenda exists to produce, and has confirmed the Responsiveness Index will be phased in as the sole eligibility gate for the trusted-neighbour badge over the coming two years.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.gjzz73.