Master of Science, Miami University, 2022, Computer Science and Software Engineering

Knowledge maps have been widely used in knowledge elicitation and representation to evaluate and guide students' learning. To improve upon current computational approaches to construct and assess knowledge maps, this thesis adopts a hybrid methodology that combines machine learning techniques and network science. By providing methods to extract features to evaluate knowledge maps and expand the assessment scope by accounting for group interaction and multiple expert maps, this thesis addresses the overall gap of current approaches for map construction and assessment. Specifically, this thesis offers three major contributions: 1) identifying necessary and sufficient graph features for knowledge maps evaluation, 2) assessing the role of group interaction during knowledge map construction and how group size affects the quality of map construction, and 3) providing an algorithmic framework to capture differences between student maps and multiple expert maps. Finally, this thesis examines the implications for the fields of network science and educational technology of applying knowledge maps in student learning.

Committee: Philippe Giabbanelli Dr. (Advisor) Subjects: Computer Science; Education

Doctor of Philosophy (PhD), Wright State University, 2016, Human Factors and Industrial/Organizational Psychology PhD

Current car navigation systems use maps that show part of a region and are sequentially presented as the driver moves along a route, displaying information that is relevant to immediate guidance, such as the surrounding streets and turn indicators. Rizzardo, Colle, McGregor, and Wylie (2013) have shown that sequentially presented, partial maps populated with landmark objects can also facilitate spatial knowledge acquisition. Spatial knowledge is useful for evaluating GPS instructions and navigating after the fact. However, the optimal number of landmarks on map segments has not been extensively tested. The Object-Based Spatial-Episodic Representations for Visual Environments (OBSERVE) theory indicates that sets of landmark object-relations are an important component of spatial learning. Landmark objects do not necessarily need to form a metric coordinate system, although sets of objects may have quantitative spatial (e.g., angular) relations among themselves, learned episodically (Colle, 2015). The number of concurrent landmark icons present on map segments was manipulated to determine the optimal range of landmark object-relations that can facilitate learning spatial knowledge of a complete region. Participants viewed a series of map segments showing a car being guided on a route with two, four, or six landmarks present on each segment, and then participants drew sketchmaps of the complete region through which the car drove. Configural spatial knowledge of the layout was measured by comparing the angular relations between pairs of landmarks on the participants' sketchmaps and the region's actual angular relations between those same pairs. Results indicate that a range of two to four landmarks per map segment that are trip-relevant is appropriate for acquiring configural spatial knowledge.

Committee: Herbert Colle Ph.D. (Advisor); Nathan Bowling Ph.D. (Committee Member); Robert Gilkey Ph.D. (Committee Member); Valerie Shalin Ph.D. (Committee Member) Subjects: Psychology

Master of Science (MS), Wright State University, 2008, Human Factors and Industrial/Organizational Psychology MS

Configural spatial knowledge has been tested by having people point from one object to another or by having them sketch maps from memory. Several different pointing judgments have been used, but these judgments appear to differ both in superficial characteristics and in their implied theoretical mental model of spatial representation. This experiment compares two different pointing judgments: judgments of relative direction, based on a quasi-Euclidean model of spatial representation; and object-based judgments, based on an object reference model of spatial representation. Results supported the object reference model. Object-based judgments were more accurate, were made with more confidence and had shorter latencies than judgments of relative direction. Analyses of the sketch maps were consistent with the pointing judgments, suggesting the results reflect stored memory representations and not retrieval differences. Issues of generality of the results and practical ramifications of the research are discussed.

Committee: Herbert Colle (Advisor) Subjects: