Читать книгу GIS Research Methods - Steven J. Steinberg - Страница 101

As a researcher, you also have the option of employing an inductive model in your research design. This type of approach begins with a different series of steps than those traditionally used for a deductive approach. An inductive approach begins with the data and proceeds to gleaning an understanding of themes and patterns. From this information, theory is then generated.

Grounded theory is an inductive research approach characterized by its sequencing: data collection followed by theory generation. It is called “grounded” because of its strong connection to the reality that the data represent. This inductive research approach is qualitative in nature. Grounded theory is an appropriate research method for assessing case studies, transcripts, oral histories, and archival data.

Glaser and Strauss (1967) first coined the term grounded theory in the late 1960s, in their seminal book, The Discovery of Grounded Theory: Strategies for Qualitative Research. Since that time, many other qualitative researchers have adopted and written about grounded theory. Grounded theory has become a popular approach embraced by a variety of disciplines, including public health, business, and criminology, just to name a few.

The key to determining if you will use grounded theory is to consider the purpose of your research. One of the primary attractions of grounded theory is that it provides the opportunity to “generate theory that will be relevant to [scientists’] research” (Glaser and Strauss 1967, vii), unlike verifying theory, which is used when following the traditional scientific method, which is deductive. Grounded theory is a good approach to employ when you are interested in the discovery phase of gathering information, because it is more appropriate for researchers whose goal is to generate information, themes, and patterns, not to prove theory.

The main premise of grounded theory is that theory emerges from an examination of the data. Rather than the researcher dictating themes and ideas that will be investigated, the data dictate what is relevant and important to study further. “Grounded Theory is based on the systematic generating of theory from data that, itself, is systematically obtained from social research” (Glaser 1978, 2). Thus, the grounded theory approach views research methods as part of the theory-generating process. The process is iterative; the researcher is constantly conducting analysis, looking for themes, and then conducting more analysis. It is a very hands-on approach to sorting through data.

The core of grounded theory is in analyzing and looking for patterns in the data. In the analysis, the researcher attempts to achieve theoretical saturation (Dey 1999). Theoretical saturation means no additional themes or concepts, categories, or relationships emerge from the data, which can only be achieved after the researcher has made a series of run-throughs with the data, identifying themes and looking for data that support the themes. Bernard (2000, 443) summarizes how grounded theory can be accomplished using the following series of steps:

1. Begin with a set of information (e.g., interviews, transcripts, or newspaper articles).

2. Identify potential themes in the data.

3. Pull data together as categories emerge.

4. Think about links between categories.

5. Construct theoretical models based on the links.

6. Present the results using exemplars.

Following these steps, you begin with whatever set of data or information you want to analyze. This information will most likely be of a qualitative nature. Identifying potential themes in your data can be done by hand or with the help of a qualitative data analysis program. As you sift through the data, certain words or phrases will begin to emerge consistently. You can then use the themes that you identify to develop a coding scheme (see Strauss and Corbin 1997) to complete the analysis of the themes. Step 3 calls for grouping, or categorizing, your information. In essence, you look for similarities, differences, and repetitions that occur in what has been stated. This is an iterative process that evolves as you analyze the data using your own specific coding process. (For more specifics on coding your data, see Dey 1999.)

Step 4 calls for thinking about the links between the grouped categories that you have seen emerge. This is akin to developing a conceptual framework or model. (See the section “Stages of sociospatial research for deductive research,” step 5.) This leads to the next natural step, which is to construct a theoretical model based on the links that you observed. This is your best model of the relationships that you saw emerge between themes that you identified in the data. Finally, in step 6, you present the data using exemplars. These are nothing more than quotations or snippets from the data that illustrate the themes, concepts, or relationships that you are discussing. You can think of exemplars as examples (shared words, quotes, etc.) of concepts or themes that emerge from the data analysis process.