The scientific method and GIS

The scientific method is a procedure through which scientific research takes place. It is used to gain insights into the observable universe through the analysis of empirical data, and to answer questions and create new ones. Due to the broad nature of geography as a discipline, as well as the fact that it is impossible to perform many geographic analyses in a controlled environment, it is necessary to take certain considerations into account when applying the scientific method to geographic data.

The scientific method process
The scientific method follows these steps:
 * 1) Observe the phenomena and create a question based on these observations.
 * 2) Create a hypothesis, or a preliminary answer to the question.
 * 3) Select and utilize any existing theories, postulates, or assumptions pertaining to this phenomena.
 * 4) Gather data and decide which data are useful in solving the problem.
 * 5) Perform experiments and record the results.
 * 6) Accept or reject the hypothesis.
 * 7) Form conclusions and new theories.

The hypothesis may be refined or rejected before all the steps are performed based on discoveries from pre-existing data and experimentation, the collection of data or materials needed to perform experiments, or as a result of the experiments.

Scientific Control
Every scientific experiment needs a scientific control in order to create a baseline of data that has not been affected by the variable being tested. It is an experiment or observation that is meant to show the phenomena in question without under "ordinary" circumstances. This control group makes it possible for scientists to compare their actual results to results unaffected by their methods.There are two different kinds of controls: positive and negative. Negative controls are control groups that weren't exposed to the treatment. Positive controls are groups that weren't exposed to the treatment, but were exposed to a different treatment that is supposed to produce similar results to the original study.

Experiment Example: studying samples on soils that were affected by certain chemicals.
 * 1) Positive Control: a sample of soils affected by different chemicals other than the ones being studied.
 * 2) Negative Control:a sample of soils that weren't exposed to the chemicals studied.

Issues in Field Science
Field science could be defined as those avenues of research that must be studied in the world, as opposed to Laboratory science that can be studied in a laboratory environment. The former would include such disciplines as wildlife biology, political science, economics, and geography, while the latter would include disciplines such as particle physics, chemistry, and microbiology. While the scientific method, with its expectation of tightly controlled experiments, works well in the lab, field science presents a number of challenges:
 * Lack of Control: it is extremely difficult to find two situations that are identical in every way except the hypothetical variable(s). In fact, the scientist may not even be aware of many of the variables involved.
 * Non-manipulability: most of the variables involved in the studied phenomenon cannot be altered by the scientist (e.g., the weather, the global economy)
 * Negative impacts: many possible experiments cannot be conducted because they could have long-lasting negative impacts on real individuals and systems.
 * Time constraints: often the processes being studied are extremely slow, and the scientist cannot wait for the effects of an experiment, especially in fields such as evolutionary biology, geology, and climate change.
 * Resource demands: conducting a significant experiment large enough to impact a real world phenomenon is often prohibitively expensive.

For example, if you were hoping to study the societal effects of hospital location, a fully controlled experiment would require that you find two towns that share the same characteristics, then build a hospital Town A and leave Town B without a hospital. Then you would spend the next several years measuring how many more people died in Town B than Town A, making sure every subsequent action in the towns was identical except for the experimental variable. This is impossible for several of the reasons above.

Modeling and Simulation
The most common solution to these issues in field science is to employing models and simulation, using tools such as Geographic information systems and Inferential Statistics. Because a model is a representation of the real world, it can be manipulated and experimented on quickly, cheaply, and with no real-world impacts. This approach is not perfect, primarily because the model is never a perfect representation of the real world situation, containing only the information the scientist chooses to represent. This can lead to a number of problems, and the results are rarely as reliable as a fully controlled laboratory experiment. However, the ability to derive theory from simulation-based science is better than most of the alternatives.

The Role of GIS
Geographic information systems can be a valuable tool in several phases of a modeling approach to the scientific method:
 * Observation and Hypothesis creation: maps and other forms of Geovisualization, as well as Exploratory data analysis, portray geographic phenomena in ways that allow researchers to quickly recognize patterns and look for possible causal processes.
 * Data gathering: the core of GIS is its ability to manage very large volumes of spatial data.
 * Experimentation: a common use of GIS in this phase is to simulate the effects of a hypothesis on a modeled situation, using techniques such as predictive models. The resultant effects of such models are then typically tested for significance using inferential Statistics.
 * Reporting: maps are one of the most useful methods for visually portraying the results of a spatial simulation.