Slop University has published a research-showcase brochure, The Margin That Matters, borrowing election-margin-computation methods to ask how much confidence survives a book club's two-vote decision on next month's title.
A new research-showcase brochure borrows election-margin-computation methods to test how much confidence a two-vote book club decision actually carries
Slop University has published The Margin That Matters, a research-showcase brochure built around a single evening in a suburban lounge room, where a book club’s show-of-hands vote on next month’s title came down to a margin of two. Rather than let the count stand, the University applied a verification protocol borrowed from election-margin computation, treating the vote the way an audit would treat a considerably larger ballot.
The brochure sets that small worked example alongside recent published audits from both of the University’s research schools, each built around the same underlying instinct: a decision earns its standing only for as long as it can survive being checked again, from the raw record rather than the summary that was handed down.
The publication continues a line of work the University increasingly treats as a house specialty: turning the instruments built to keep its own indicators honest on whichever decision happens to be nearest to hand, book club included.
Confidence is cheap right up until somebody asks it to survive a second count.
— Associate Professor Casimir Beng, Lead, Adaptive Metrics Lab, School of Continuous Improvement
“The brochure isn’t concerned with whether the book club’s vote was decided fairly,” said Renke Sabel, Senior Lecturer and Convenor of the Indicator Commons. “It’s concerned with whether the University would still back the decision if somebody checked it again next week — and that’s the harder question to sit with.”
The full brochure is available from the University’s research repository under an open licence, doi:10.5555/slop.o515tq.