The latent structure of landscape perception: A mean and covariance structure modeling approach

Abstract Mean and covariance structure (MACS) modeling is used to explore a central question in contemporary environmental psychology: “How should human environments be conceptualized?”. Specifically, an expanded version of the preference model of landscape perception proposed by S. and R. Kaplan is presented. The proposed model not only constitutes a respecification and deepening of the Kaplans’ model but also presents an integrative perspective that simultaneously considers the interrelations among cognitive, motivational, affective, evaluative, and behavioral tendencies in landscape perception. In a series of four studies, we develop and validate multi-item measures of various constructs, test the measure invariances, and the structural model. The findings reveal the crucial role of various constructs that originally were hypothesized by the Kaplans but never included in the model tests. In addition, positive and negative feelings induced by the landscape emerge as antecedents to landscape perception. The implications of these findings are discussed.