Searching for Legibility

Abstract

Legibility has been ineffective as a predictor of environmental preference primarily because of its correlation with another predictor, coherence. The authors tried to separate the two predictors by careful selection of field/forest settings and by using nontraditional definitions. The alternate definitions emphasized landmarks (for legibility) and the two-dimensional picture plane (for coherence). These strategies proved unsuccessful for the entire sample of settings. However, when an empirically derived subset of forest settings was examined, the desired pattern of relations among the traditionally defined constructs was found: Legibility had a slightly stronger correlation with preference than coherence, and legibility was clearly the stronger predictor in regression models that included several predictors. Post hoc analyses involving openness suggested visual access is a major component of legibility in forest settings. The authors now believe the forest setting category is a good domain for establishing the salience of legibility as an independent predictor of preference.