A School of Emergent Priorities paper taxonomises shift workers' timesheet-correction requests at a national cleaning contractor, finding one reason category predicts an audit reversal, and the contractor turning the finding into a filing-rights policy within six weeks.
A five-category taxonomy of cleaning-contractor timesheet corrections finds one reason predicting an audit reversal, now written into policy
Dr Ingrid Vasseur convenes Futures in Committee, the School of Emergent Priorities’ seminar series on how a scanning exercise’s open questions narrow into an agenda item and a minute. Her latest paper, written with colleagues from the School of Continuous Improvement, finds a plainer version of the same narrowing at work somewhere nobody had thought to look for it: the reason field of a timesheet-correction request.
The team taxonomised 2,918 correction requests filed over one payroll quarter by shift workers at a national cleaning-services contractor, sorting each into one of five reason categories drawn from the requests themselves. One category stood out: requests filed with no stated reason at all were reversed at the contractor’s next payroll-integrity audit at close to five times the rate of the other four combined. The finding sits squarely inside the School’s wider account of how a scanning question hardens into standing business once an institution has somewhere to file it.
A field this quiet about itself turns out to be the one worth reading closest.
— Dr Ingrid Vasseur, Lecturer and Convenor, Futures in Committee
Within six weeks of the finding reaching the contractor, its payroll-integrity office had retired the free-text field, adopted the five-category taxonomy as mandatory, and introduced a rule suspending a worker’s correction-filing rights for a quarter after a third blank-reason-coded request. Dr Sten Okwuosa, Senior Research Fellow and Convenor of the Impact Pathway Atlas, said the six weeks since “are the part of the record we keep returning to, less for what the audit numbers did than for what the category itself quietly became.”
The team’s approach borrowed the coding discipline of qualitative content analysis, cross-checked between two independent readers, and paired it with the kind of before-and-after comparison the School increasingly reaches for once a finding is handed back to the institution that produced the data behind it.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.bmkewc.