A School of Continuous Improvement paper reports a pre-registered replication of an earlier bin-night roster finding across 31 streets, finds no effect, and reports the Review Cadence Observatory fixing every audited council's roster reissue to a quarterly calendar regardless of the null.
A 31-street replication of an earlier roster-decay finding turns up nothing, and every audited council moves to a fixed quarterly reissue regardless
A quasi-experiment that has anchored a month of the University’s own citation graph — the finding that a laminated bin-night roster loses its grip as it ages, with households drifting into copying whichever bin a neighbour has already kerbed — has failed to reproduce in a pre-registered re-audit of 31 streets in a different council portfolio, the School of Continuous Improvement reports.
The re-audit, run by the School’s Review Cadence Observatory, matched the original study’s photographic method street for street and found no relationship between roster age and either outcome: not the original decay, not the neighbour-copying rise, and not even a smaller version of either surviving inside the original study’s own council area.
A finding doesn’t stop being ours just because the second try didn’t find it. We would rather publish the second try.
— Dr Dagny Okafor, Lecturer and Convenor, Evaluation of Evaluation, School of Continuous Improvement
Dr Osei Vandermeer, Research Fellow and Convenor of the Review Cadence Observatory, said the absence of an effect was, if anything, the easier result to act on: a schedule that doesn’t depend on the roster actually decaying, he said, “is one we can keep no matter what the next audit finds.”
On the strength of the non-result, the Observatory has recommended every council in its audit portfolio move to a fixed quarterly reissue of its bin-night roster, independent of any street’s measured compliance — a rule its own proposal describes as the more defensible one to keep precisely because it no longer needs the original finding to be true.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.lhmgn3.