Publication Date


Document Type


Committee Members

Herbert Colle (Advisor), Assaf Harel (Committee Member), Scott Watamaniuk (Committee Member)

Degree Name

Master of Science (MS)


Spatial memory for the layout of large-scale environments, configural spatial memory, has typically been construed as being very structured, using something like a metric coordinate system and using environmental objects to define that coordinate system. Inside of buildings, rectangular rooms have walls at right angles that have been considered to fulfill this role. However, the influence of non-spatial factors and considerations of relatively unstructured environments have not received much attention. Semantic organization was found to improve configural spatial memory for landmark objects in rooms with walls and it was independent of the structural relations among landmark objects (Colle & Reid, 2000). The mechanism behind this semantic effect is not well-understood. The present study also used semantic organization (grouping landmarks) and manipulated structural information in a different way, by comparing walled rooms with equivalent non-walled quadrants. It also randomized landmark object placement, providing minimal structural cues in non-walled conditions. Participants experienced a single tour of four rooms/quadrants using a random path to visit each landmark object. Participant performance was measured by having them both create sketchmaps of the environment and make angular judgments between objects using a direction circle. As expected, absolute angular error was smaller for walled environments than those without walls. Results from the sketchmaps showed that semantically grouped landmarks improved performance when walls were present, but the effect was not statistically significant without walls. In contrast, results from directional pointing queries, the other memory retrieval measure, showed that semantically grouped landmarks did improve performance without walls, but the effect was not significant when walls were present. These data suggest that people can acquire configural spatial knowledge quickly in relatively unstructured environments and that verbal effects can improve spatial memory in both structured and relatively unstructured conditions. Potential explanations are discussed.

Page Count


Department or Program

Department of Psychology

Year Degree Awarded