Steven J. Steinberg

GIS Research Methods


Скачать книгу

Learning objectives

      

Comprehend the challenges facing geography, information, and system aspects of spatial research

      

Discover trends in the data

      

Learn the phases of abstraction and how they relate to research

      

Learn the differences between a logical data model and a conceptual data model

      

Understand the importance of data aggregation to spatial research

      

Learn what questions to ask about geographic data, location, and analysis

       Key concepts

      basemap

      cartograms

      code data

      computer animation

      conceptual model

      coordinates

      data aggregation

      data dictionary

      datums

      digital

      fuzzy GIS

      index

      key informant

      latitude

      logical data model

      longitude

      oral history

      phases of abstraction

      scale

      social networks

      trends in the data

      variability

      visualizing data

      GIS are best understood by breaking down the terminology and learning how to apply GIS to various analyses. In particular, how can your area of interest and the associated data be placed into a GIS context, and how can GIS technology enhance your analysis and understanding of data? Here we review GIS in detail, letter by letter, to establish this understanding, before discussing data conceptualization.

       The G in GIS

      The geographic component of GIS is simultaneously obvious, confusing, and difficult to master. From an early age, we all develop an understanding that the locations of people and places can be marked on a map, and furthermore, that connections can be made between these locations. What we may not have is a good understanding about the scientific basis for mapping—that is, the numerous issues of scale, coordinates, control datums, and so forth. Other than mapping professionals, very few people have a deep conceptual understanding of the mathematical algorithms behind these concepts and the potential errors that result from various combinations and interactions of such data.

      Fortunately, most of these underlying issues are addressed for us through the GIS software, so it is not essential to have a deep understanding of them. It is important, though, that you pay attention to a few essential concepts, even if you do not know exactly how they work. You can think of this as analogous to knowing the difference between a CD player and an MP3 player—you know these are different tools with different strengths and weaknesses, but selecting the right one does not require you to understand their inner workings. What is important is that you know which format to ask for so that the medium selected fits the player you own.

      GIS has been used to locate and manage natural resources and is a well-embraced technology among many in business and marketing. Although GIS is very valuable to social science research, it has not been incorporated as frequently into this field. Many social science studies focus on social, economic, cultural, and survey data, asking questions such as, Do pregnant women who are better educated or wealthier receive higher-quality prenatal care? Perhaps more probing questions would ask about the locations of prenatal clinics relative to available public transportation, child care, and so on. You might use census data to conduct a statistical analysis of census block groups and levels of prenatal care but may miss an important locational component. Often, when explored in conjunction with other map-based location information (e.g., where are the blocks located relative to other important components?), a more complete understanding of the causative relationships can be obtained. Furthermore, from an applied standpoint, the geographic component can help in determining where to best locate and spend limited resources to help improve the situation.

      In reality, almost all information researchers collect about people, their communities, and their environments can be tied to some geographic location. For example, you may survey people at their home address or by some geographic unit such as a census block or city of residence. All of these locations can be easily mapped. Furthermore, if privacy is a concern, you can engage in data aggregation, which means using a larger geographic unit to mask specific, personal information. In short, if you can answer the question, Where were the data collected?, then GIS is an appropriate means for storing and analyzing the data.

       Difficulties with the G

      The geographic context may be difficult to collect because determining the exact location of a piece of data on the ground is not always easy to accomplish or, for reasons of privacy, may not be permissible. When mapping people, we face an additional challenge: people may move around, may be without a home, or otherwise may be difficult to tie to a particular location. However, because geographic data are the heart of GIS, knowing a location of some kind is an essential part of the GIS process (even if it must be spatially degraded or detached from the exact, true location).

      For example, if you were doing a study of homeless individuals, it might be better to define their location at the level of a particular neighborhood they call home than at a particular street address. Furthermore, even in studies where mappable locations are available, privacy issues may necessitate degrading that information. In other words, even if you have specific addresses of your respondents, you might choose to degrade the data to census blocks, to neighborhoods, or even to the city level to maintain the privacy required for ethical research. Choosing the level of spatial detail is an important part of the GIS process.

      Conceptually mapped features, such as data about perceptions, ideas, or interactions (figure 2.1) are perhaps more difficult to map; although, they are equally important as physical locations in research. For example, social networks or interactions between individuals may be mapped in such a way that people who are emotionally close would be located conceptually close together, whereas individuals who are casual acquaintances might be mapped at a greater distance. Lines connecting people on the map could represent social distance rather than true geographic distance. On such maps, referred to as cartograms, the distance between mapped data is scaled to a variable or index value other than distance. In the case of social ties, this might be an index representing the strength of a particular relationship.

      Figure 2.1 University of the Arctic thematic networks map. This map presents collaboration networks between higher education and research institutions in the northern hemisphere. The map’s goal is to illustrate the networks of activity and