Abstract
The process with which the visual quality of our rural landscapes is classified, assessed, and planned frequently utilizes public-participation data. This data can be quantified for use in a planning tool. This thesis was designed to obtain data on the way that observers/users, both trained and untrained alike, visually classify landscape scenes into categories, and also what visual preferences are for those categories. For classification, a q-sort (non-evaluative, unforced distribution) was administered to a small group of trained observers (n=9), and a factor analysis was used to derive categories from data collected from a visual preference survey (n=70). Cluster analysis was used to analyze the q-sort results, and these were "layered" onto the factor analysis results which resulted in a final set of 3 main landscape categories. Statistical tests were run to determine which landscapes and landscape categories were the most preferred, and who preferred them. The results indicate that those with a more urban back ground tend to have higher preferences for the rural landscape, education may positively correlate with preference for agricultural landscapes, and that age and length of residence may positively correlate with preference for a town/village setting.