The use of design elements to predict visual compatibility between natural and built environments, scenic beauty, and severity of impact

TR Number



Journal Title

Journal ISSN

Volume Title


Virginia Polytechnic Institute and State University


Two studies were conducted to determine the visual elements that influence visual compatibility between natural and built environments. In the first study subjects (N=109) rated one of three versions of color slides of 20 western United States landscape scenes on how much they consisted of 11 simple visual elements (Round, Square, Conical, Cylindrical, Vertical, Horizontal, Diagonal, Curvey, Smooth, Rough, and Relative Size). The first version was a Pre-impact scene without human-made structures. The second version was a Post-impact version of the same scene but with a structure on it. The third version was a Structure-alone version, consisting of only the structure. Subjects rated either the Pre-impact or the Structure-alone scenes on the 11 design element dimensions. For each scene the absolute value of the difference between the mean pre-impact rating and the mean structure rating on each dimension were the measures of compatibility between the structures and their environments to be used as predictors in subsequent regression analyses. Independent groups of subjects rated the Pre-impact scenes on scenic beauty, while others rated the post-impact scenes on scenic beauty, overall compatibility, and severity of impact (of the structure). These mean values for each scene provided the criteria values for the regression analyses. In general, the results showed that the compatibility measures of Rough texture and Relative Size were strong predictors of the criteria. Diagonal and Conical element compatibility were less pervasive but significant predictors, as well. The second study (N=l36) was similar to the first but used 45 simple line drawings of and Square and landscapes. Vertical In this study Relative element compatibility Size, were significant predictors of the criteria, but accounted for less variance than the study one models. Some predictors (e.g., Square elements) predicted in opposite directions across the two studies.