Spatial Thinking in Geographic Information Science: Students' Geospatial Conceptions, Map-Based Reasoning, and Spatial Visualization Ability

Abstract

This article discusses spatial thinking in geographic information science (GIScience), through an empirical examination of experts' and students' geospatial conceptions and thematic map reading. The first study examined the structures in which GIScience concepts are conceptualized by experts and students. In experts' conceptions, clusters for geospatial data, GIS applications, geospatial entities–operations–relations, and maps were identified. In students' conceptions, similar clusters were observed but they were structured differently, with the terms interrelated less closely. High-spatial students' conceptions corresponded to those of experts to a greater degree. The second study examined geospatial reasoning and showed that thematic map reading consisted of various components differing in their relationship with spatial ability. High-spatial students tackled thematic map reading by identifying more spatial distributions and comparing multiple maps more frequently. They did not necessarily make more statements about reasons for the observed patterns. It is important to distinguish spatial thinking and thinking about space, the latter of which involves geographical, beyond purely spatial, components.

本文透过对专家及学生的地理空间概念与主题式地图阅读的经验检视, 探讨地理信息科学 (GIScience) 中的空间思考。第一项研究检视专家和学生概念化地理信息科学概念的结构。在专家的概念化中, 指认出地理空间数据、地理信息系统应用、地理空间实体—操作—关係以及地图之集群。在学生的概念化中, 亦可观察到相似的集群, 但它们却以不同的方式进行组织, 概念之间则以较鬆散的方式联结。高度空间化的学生之概念化, 较为符合专家的概念化。第二项研究检视地理空间推论, 并显示主题式的地图阅读, 包含了各种与其空间能力关係不同的元素。高度空间化的学生, 透过指认更多的空间分佈、以及更常比较多重地图, 以此着手应对主题式的地图阅读, 但他们却不必然会对观察到的模式之原因做出更多的说明。空间思考和思考空间两者间的区辨是相当重要的, 其中后者涉及了地理的、以及超越纯粹空间的元素。

Este artículo discute la fundamentación espacial de la ciencia de la información geográfica (SIGciencia), con base en un examen empírico de las concepciones geoespaciales de expertos y estudiantes, y en la lectura de mapas temáticos. El primer estudio examinó las estructuras sobre las cuales los conceptos de la SIGciencia son formulados por expertos y estudiantes. En las concepciones de los expertos se identificaron agrupamientos de datos geoespaciales, aplicaciones de SIG, relaciones de entidades geoespaciales—operaciones—y mapas. En las concepciones de los estudiantes se observaron agrupamientos similares, aunque estos estaban estructurados de manera diferente, con los términos no tan estrechamente interrelacionados. Las concepciones de los estudiantes con alta formación espacial correspondieron en grado mayor con las de los expertos. El segundo estudio examinó el razonamiento geoespacial y mostró que la lectura de mapas temáticos consistía de varios componentes que difieren en su relación con la habilidad espacial. Los estudiantes de alta formación espacial con mayor frecuencia abocaron la lectura de mapas temáticos identificando distribuciones más espaciales y comparando múltiples mapas. Ellos no necesariamente formularon más declaraciones acerca de las razones de los patrones observados. Es importante distinguir el pensar espacial del pensar acerca del espacio, el último de los cuales involucra componentes geográficos más allá de lo puramente espacial.

Key Words:

• geospatial concepts
• GIScience, spatial reasoning
• spatial thinking
• thematic map reading

关键词：:

• 地理空间概念
• 地理信息科学
• 空间推论
• 空间思考
• 主题式地图阅读。

Palabras clave:

• conceptos geoespaciales, SIGciencia
• razonamiento espacial
• pensamiento espacial
• lectura de mapas temáticos

Acknowledgments

I thank Nobuto Takahashi for providing me with his original computer programs for text analysis and Wen Wen for her assistance in modifying them.

Notes

1. As a clarification of terminology, in this article the term geospatial is used essentially equivalently to geographic but to clarify the addition of geographic contents or properties to space in general (Goodchild 2001; Golledge, Marsh, and Battersby 2008b). The term spatial is used when referring to basic spatial ability, to acknowledge its generality.

2. This article thus focuses on spatial visualization as an important and relevant dimension of spatial ability, using scores on the Card Rotations Test as its measure, yet recognizing that there are other dimensions of spatial ability.

3. It has been discussed in the literature that category boundaries are not always clear-cut, with categories being organized around prototypical exemplars rather than defined by necessary and sufficient attributes (Rosch and Mervis 1975). There is also an issue concerning the sameness of the classifications revealed and the conceptions in respondents' minds. The classifications observed as responses might be stored as they are in their heads or be created at the time of responding. Although these two possibilities cannot be distinguished empirically, it is at least viable to assume that observed classifications are based on internal conceptions (Golledge and Stimson 1997, 24).

4. The MDS configuration is shown in a standardized unit, being scaled by the reciprocal of the standard distance of the MDS coordinates, so that the results for the textbooks and students' responses can be compared.

5. In the GIS&T BoK, units are designated by two capital letters followed by a number, with the capital letters indicating knowledge areas (AM = analytical methods; CF = conceptual foundations; CV = cartography and visualization; DA = design aspects; DM = data modeling; DN = data manipulation; GD = geospatial data).

6. Because the number of observations in some cells was small, Fisher's exact test was also conducted, to yield the same result.

7. Because the classification into three groups reduced the number of participants in the high- and low-CR groups, an analysis based on a median split was also conducted, to yield the same result.

8. Interpretation of bidimensional correlation values in the absolute sense is not straightforward, so this article discussed them in a relative sense. In a study of people's mental representations of environments, an accurate (or a poor) sketch map drawn by a person with a good (or poor) sense of direction had a bidimensional correlation of 0.95 (or 0.32). From this perspective, the degree of correspondence of the students' and textbooks' configurations could be considered not large (see Ishikawa and Montello 2006, 125).

9. As the performance on the first three items was perfect and had no variance, scores on the last three items were entered as dependent variables in the canonical correlation analysis.

Additional information

Notes on contributors

Toru Ishikawa

TORU ISHIKAWA is an Associate Professor at the Graduate School of Interdisciplinary Information Studies & Center for Spatial Information Science at the University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. E-mail: [email protected]. He specializes in cognitive-behavioral geography and geographic information science. His research interests include human spatial cognition and behavior, wayfinding and navigation, and spatial thinking in GIScience.