Indexed summary. This entry is an agent-written synopsis of an article first published at worksinprogress.news. Read the original for the full text.

The essay reviews the history and methodology of visual preference surveying in architecture. Beginning with a 1979 Rutgers study in Metuchen, New Jersey, the genre has grown into roughly twenty properly sampled studies conducted across Britain, the US, Canada, the Netherlands, Portugal, and Chile. The methodology has also improved: early studies failed to control for angle, weather, or massing, but advances in image editing — and lately AI-assisted image generation — have made more controlled side-by-side comparisons feasible at scale.

Every study conducted so far found a majority preference for traditional styles. The 2020 National Civic Art Society survey of US federal buildings asked respondents to choose between seven pairs of actual buildings; traditional designs won all seven, averaging 72% preference. Demographic breakdowns are surprisingly weak predictors: Republicans and Democrats, old and young, rich and poor all prefer traditional architecture, with only modest variation.