Читать книгу GIS Research Methods - Steven J. Steinberg - Страница 96

As you consider the selection of your research methods, it is important to refer back to your conceptual framework. You need to decide exactly what variables you will measure, and how. This requires that you define each of the concepts that you are using in your study and its associated level of measurement. Operationalizing a variable requires you to explain how you are going to measure the concepts mentioned in your hypothesis. Typically, the variety of data collection approaches that you employ is documented in the methods section of your study. When your methods relate directly to the creation of a GIS dataset, you need to be sure to record the data creation procedures in a metadata file, which should be included with each individual data layer.

In the example of the relationship between pollution and poverty, you might choose to operationalize pollution as shown in figure 3.9 into three simple categories: low, medium, and high. In doing so, you would need to define exactly what is meant by these terms. Doing this would be an example of translating data from a simple numerical form into a perhaps more palatable and easily understood frame of low, medium, and high. Of course, you would have to explain the data that are used as part of your categorization process for low, medium, and high. For instance, is “low pollution” something that is measured as a number of pounds or gallons released, or by a medical definition such as the odds of getting sick from exposure (e.g., one in 1 million as low; one in one hundred thousand as medium; and one in ten thousand as high)? Of course, you could come up with a variety of definitions, each viewed as appropriate or inappropriate, depending, for example, on legal definitions, regulatory recommendations, expert opinions, or local experience.

Coming up with a true definition is not necessarily as important (or possible) as explaining your definition so others can understand and interpret your results. One advantage of the GIS is that if you choose to change your definition or examine different scenarios, it is relatively easy to rerun your analysis with the adjusted definitions to determine how the results change.