A new research poster from the School of Continuous Improvement finds a jellybean-jar guessing competition's open sign-in sheet converging tightly on itself across a stall session, even as its guesses drift further from the jar's counted total.
An 11-week Union Market study finds a visible sign-in sheet's guesses grow more alike without growing more accurate
The Union Market’s Jellybean Census has run for three consecutive terms on a simple pairing: a jar, and either a sealed ballot box or a pinned sign-in sheet where each guess sits in view of the next. A new research poster from the School of Continuous Improvement, Converging on Each Other, Not the Jar, is the first to check what the open sheet’s famously tightening spread of guesses has actually been converging on.
The finding lands as the School’s evaluation-ecosystem agenda turns, this term, toward the instruments campus events run on without a second look. Researchers logged 428 guesses across eleven stalls over the full term, split between the sealed and open conditions, matching each stall’s spread of guesses against the jar’s own count once judging closed. The open sheet’s spread narrowed sharply as each stall’s queue moved through it; the sealed box’s did not — and the open sheet’s accuracy against the counted jar did not improve to match.
The sheet was never measuring the jar. It was measuring how quickly a queue agrees with whoever’s already written something down.
— Dr Mirela Hanke, Postdoctoral Fellow, School of Continuous Improvement
Dr Sten Okwuosa, Senior Research Fellow and Convenor of the Impact Pathway Atlas, said the result “raises the obvious next question for every board on campus that shows its working as it fills up” — namely, which ones are converging on an answer, and which are just converging.
The full poster is available from the University’s research repository under an open licence, doi:10.5555/slop.hh9kvt.