School of Emergent Priorities researchers instrumented the naming of shared workplace and club machines on the pronounceability of their model designations, and report that a naive association worth nearly four years of extra service does not survive identification.
An instrumental-variables study of 1,806 items at 214 sites, and the 1,412 equipment names the Horizon Register has since removed
Somewhere around half the machines a small workplace or community club shares have a human name, and the people who use them will tell you the named ones last longer. A research poster released today by the School of Emergent Priorities set out to establish whether the name is doing any of that work.
Naming a Shared Machine Does Not Extend Its Service Life draws on asset-register extracts, service dockets and acquisition invoices for 1,806 items held at 214 workplaces and community clubs. The naive comparison is striking: a named item stays in service some 3.6 years longer than an unnamed one. Once naming is instrumented on how pronounceable each item’s model designation happens to be, the effect falls to 0.2 years, with a confidence interval wide enough to accommodate almost nothing. No class of equipment in the study retains a positive effect at conventional levels.
A result that sends you back to the register rather than out into the world is still a result. We know rather more now about what our own records were carrying, and rather less about why.
— Dr Fenna Okoro, Senior Lecturer and Convenor, Horizon Register
Associate Professor Kwame Lindqvist, who joined the team from the School’s work on commitment durability, said that identification was the whole of the exercise: “The association was never in doubt. What it meant was.”
The work continues the School’s interest in what a record keeps after the thing recorded has moved on, and the University welcomes the questions it raises about attributes it stores without ever testing. The Horizon Register has since removed 1,412 recorded equipment names as unverified attributes, reporting the removals as records rectified.
The full poster is available from the University’s research repository under an open licence, doi:10.5555/slop.ioerp9.