Whether data used to produce a map is entered manually into a computer program such as Microsoft Excel or Access, collected via Global Positioning System (GPS) units then transferred to mapping software, or collected manually by some other means, the first step in producing a computer or GIS-based map is to obtain data. What a researcher wishes to display on a map will determine how the data will be obtained and entered into a GIS.
For example, if a researcher is studying how local residents feel about a new store coming to town, data will not likely exist before the beginning of the study. The researcher would need to collect the data manually and then enter it into a GIS data table. On the other hand, if a researcher wishes to study the patterns of tornado distribution in the United States, data that have already been collected by the National Weather Service can be downloaded and imported directly into a GIS. The availability or lack of data determines how data will be entered into a GIS.
Researchers using GIS have the ability to display any information about the available data and create maps specific and appropriate for that information. Within a GIS environment, the characteristics (referred to as attributes) that describe the data can be displayed as point symbols which can be scaled to reflect the relative magnitude of some phenomenon; this method is referred to as graduated symbols. A researcher can also display data by methods such as dot density, where one dot is equal to some predetermined value, or by a color scheme, where each color on the map reflects some aspect of the data. The use of computer software in the form of a GIS allows for easy manipulation of data to determine the most suitable form for display.
Many statistical techniques have been adapted to GIS software packages, allowing researchers to perform complex spatial analyses on large spatial datasets. Within a GIS environment, packaged scripts accept input concerning what specific test or analysis to perform, what data to perform it on, and the specific conditions of the test or analysis. For example, to learn about patterns of corn production and where the most concentrated corn-producing counties in the U.S. are located, a researcher may use a test call the Getis-Ord statistic. This analysis involves many complex mathematical steps and becomes extremely lengthy and prone to user-error if the dataset is large and the analysis is done by hand. Within a GIS, a researcher can apply the statistic to the data using the appropriate script and analyze the results in seconds. The speed of modern computer processors makes these large and complex tasks much simpler.
Researchers often collaborate on projects, and as such, it becomes extremely important to share data and to ensure that data are uniform. GIS software packages have the ability to store many files of related data in what are called geodatabases, or GDBs. These collections of data can be transferred as a single file while maintaining all of the spatial relationships between the data files as well as integrity and behavioral rules as to how the data acts. This process saves researchers time in that as a GDB is moved from one computer to another, the new user does not have to untangle another user's method of defining data; the GDB accomplishes that task.
Before computing technology became widely available, maps were made by hand. Cartographers would assemble a collection of data drawn the area on which the data was to be imposed and fill in the area or areas with appropriate data symbols, labels, scale bar and legend. Final versions of maps could be copied and widely distributed, but the original production process could be time consuming. Once a map was completed, it was difficult to make changes to the original design without starting all over. With the advent of GIS, maps can be custom-produced quickly, printed on large-scale printers called plotters or sent as digital files to anywhere in the world. The same map can be viewed in a GIS, altered if desired and printed again. Changing the entire appearance of a map to produce a unique map within a GIS environment can be accomplished relatively simply.