A new Slop University paper models 4,918 concealed-label try-on judgements over 246 garments and finds the retailer nearly as predictive of fit as the size printed on the tag.
A concealed-label try-on panel across nine retailers puts the shop a garment came from almost level with the number printed on it
A 12 from one shop and a 12 from another are, on the evidence of 4,918 try-ons, further apart than a 12 and a 14 from the same shop.
Twelve Means Nine Things, published today by the School of Emergent Priorities, treats the printed size as an instrument and estimates its parameters the way a testing agency would. A panel of 118 wearers judged 246 garments labelled 8 to 18 from nine retailers, and an item-response model placed every garment on one latent scale. The label carries 34.1 per cent of the variation in who a garment fits. The retailer carries 28.6.
A ladder that is honest inside one shop is not a standard across nine, and nothing on the garment tells you which of the two you are holding. The number is doing one job properly and a second job it was never built for.
— Professor Verity Marris, Director, Trajectory Analytics Group
The garments were bought at full price and never altered, and every label was covered before a wearer saw it. “Nobody in the panel was ever told a size, including their own,” said Dr Anneke Tolan, Senior Research Fellow. “Once the number is visible you are measuring what people expect rather than what fits.”
The paper closes on the reporting layer above the rack. The sector’s size availability index counts labels present at audit, so a retailer whose 12 sits two steps from everybody else’s scores exactly as one whose 12 sits with them, and returns attributed to fit are largely coded to a change-of-mind category that quality reporting sets aside. The University puts the work forward as its measurement programme reaching a surface most people handle weekly.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.iv44yy.