A School of Continuous Improvement instrument scoring kitchen-bench, yard and bin-night readiness across 118 households for five years found no relationship to any outcome it was meant to predict, and the School has entered it on its public indicator-retirement ledger as closed.
A composite kitchen-bench, yard and bin-night score tracked 118 households for five years, predicted nothing beyond itself, and has been formally retired
Most instruments an institution builds outlive their evidence considerably longer than they outlive their funding. The School of Continuous Improvement’s Impact Report 2021–2026, released today, tells the story of one that didn’t: a five-year household score retired the same reporting period its own evaluation confirmed it had nothing left to say.
The instrument, CHORE, scored 118 households nightly on three observable states — the kitchen bench cleared, the yard mown and edged, and the correct bin at the kerb the night before collection — rolling up into a five-year composite the School published back to participating households as a private benchmark. An independent, separately logged measure of household friction incidents was tracked alongside it throughout.
An indicator earns its keep by predicting something. CHORE never did, and the honest response was to say so on the record rather than let the number keep running out of habit.
— Dr Renke Sabel, Senior Lecturer and Convenor, Indicator Commons, School of Continuous Improvement
Five years of nightly scoring against that independent measure returned, in the University’s own description, no relationship worth reporting as one. Rather than fold the result into a call for a longer trial, the School’s Indicator Commons ran the instrument through its published retirement protocol and entered CHORE on the public ledger tagged “retired: no predictive relationship found” — the ledger’s first entry to carry that exact wording.
Dr Marit Osayande, who convenes the School’s Living Dashboard initiative, said the finding echoed her own program’s work on how quickly a live indicator stops being read once it stops being new, and welcomed the chance to test that pattern on the School’s own instrument for once.
The full impact report is available from the University’s research repository under an open licence, doi:10.5555/slop.s1askf.