A new Slop University research poster interviews 64 nominal human sponsors across 18 junior netball clubs and finds genuine oversight of an automated team-selection app grows rarer, not more common, at the clubs that lean on it most.
An interview study of 64 nominal sponsors across 18 netball clubs finds real oversight of an automated team-selection app is rarest at the clubs most reliant on it
Every automated team-selection app used by a junior netball club needs, on paper, a human sponsor: someone named against the account who is nominally responsible for what the algorithm decides. A new Slop University research poster asks what that person actually does with the role, and finds the honest answer is often not much.
The work adds to the School of Continuous Improvement’s growing catalogue of nominal human roles sitting over automated systems, tracing what each one turns out to hold once someone actually asks.
Lecturer Marit Osayande and Senior Lecturer Renke Sabel interviewed 64 sponsors across 18 suburban clubs, sorting each account into one of four categories built to span the distance between rubber-stamping a team sheet and genuinely contesting it.
A sponsor’s name on the account was never proof anyone was reading the account.
— Lecturer Marit Osayande, Convenor of the Living Dashboard
Two categories, Rubber Stamp and Retroactive Justifier, accounted for most sponsors interviewed, and the category built to capture genuine override was the rarest of the four at clubs of every size, most markedly at the largest ones.
“The clubs leaning hardest on the app were the ones least likely to have anyone reading past its summary screen,” said Senior Lecturer Renke Sabel, Convenor of the Indicator Commons. “Scale seems to compress the role rather than share it out.”
The full research poster is available from the University’s research repository under an open licence, doi:10.5555/slop.msvq20.