A Stated-Preference Experiment on Park Drinking-Fountain Attributes, a Revealed-Detour Validation at 96 Fountains, and the Siting Index Between Them
A paper from the School of Emergent Priorities re-estimates the discrete choice experiment behind an open-space amenity index six adjoining metropolitan councils use to rank candidate drinking-fountain sites, and validates it against observed behaviour at the fountains it governs. A choice survey of 1,842 residents (22,104 choice observations, mixed logit with correlated random parameters) returns a marginal willingness-to-walk of 331 m for a bottle filler, against a revealed marginal detour of 10.8 m recovered from 24,106 passers-by and 2,418 drinking events observed at 96 fountains over 328 sessions, a ratio of 30.6. Four pre-registered estimation variants move the stated figure by no more than 8 m, and the attribute ordering is preserved across both measures; re-calibrating the index weights to the revealed detours moves 14 of 96 candidate sites across the funding cut-off. A paired delivery test finds the fountain non-delivering at 31.2 per cent of visits, a condition the attribute set excludes because the design brief scoped it to features specifiable at procurement.
| Authors | Ronja Oyelaran, Fenna Okoro, Bram Ntuli |
|---|---|
| School | School of Emergent Priorities |
| Output type | Research paper |
| Funding | Anticipatory Capability Fund: Retiming the Oval's Sunscreen Restock From the Empty Bottle to the Forecast ($132,916) |
| Published | |
| DOI | 10.5555/slop.r5xph8 |
| Pages | 6 |
| Version | 1.0 |
| Licence | CC BY 4.0 |
Cite as
@misc{slop_r5xph8,
author = {Ronja Oyelaran and Fenna Okoro and Bram Ntuli},
title = {They Said They Would Walk: A Stated-Preference Experiment on Park Drinking-Fountain Attributes, a Revealed-Detour Validation at 96 Fountains, and the Siting Index Between Them},
year = {2026},
publisher = {Slop University},
doi = {10.5555/slop.r5xph8},
url = {https://slop.university/outputs/slop-paper-park-drinking-fountains-r5xph8/},
version = {1.0},
note = {Research paper},
}