While mapping is one viable tool that you can use to represent your data in a visually interesting way, there are other methods of visualization that can be used to show different relationships between points in a dataset. One way to show these relationships is with a network analysis. A network analysis shows a specific relationship between two points of data in the dataset, and visually represents this relationship in a way that is not possible with a basic map. If a connection exists between two of the points, as I will show later using a dataset based on interpersonal aid between people during the holocaust, it will create a link between the two points of data. If there is enough data in the set, with enough connections, the network analysis will resemble a web, showing the differing degrees of connections between many different data points.
Like creating a map, a network analysis can be created using both Google Fusion Tables and Palladio. Each have their own specific strengths and weaknesses, but each offer the ability to show your data in this way. For my example, I will be using a dataset, that shows who gave aid, who received aid, and what type of aid it was during a specific set of years during the holocaust, which can be found here.
The image on the left is a network analysis of this data showing the basic network analysis using basic parameters. The parameters are those who received the aid, notice that this field is represented by bolded points, and those who gave the aid, the lighter points. This data shows that there was a very intricate connected web of people all giving aid within a relatively small group.
If you want to show a different representation of the same data, for example the gender of those who gave aid, you simple have to change the target data point to reflect the field you want to show. The network on the right reflects this data target, the bolded points represent the gender of the providers of the aid, and we can tell from this that the majority of the providers were male, which is shown by the clustering around the “1” point.
Another option that there is to make a network analysis such as this is Google Fusion Tables. As the example below shows, the basic visualization of the data looks largely the same as Palladio, the main difference being the ability to add colors to the data points rather than bolding them. However, Palladio offers a very important option to the user that Google Fusion Tables does not, the ability to import a second sheet of data into the current project.
Because Google Fusion Tables does not allow the user to do this, you, at least for this specific dataset, cannot accurately scale the data point sizes based on the amount of aid received, as Google Fusion Tables does not have access to all of the information. A potential way around this would be to keep all of the necessary data on a single sheet, alleviating the issue of needing both when it only allows the upload of a single one.
One argument against using this type of visualization is the fact that the ease of use allows for many datasets to be mapped this way. In an article titled Demystifying Networks, Scott Weingart alludes to this fact by stating that almost any dataset can be visualized using network analysis, however, that does not, in his opinion, mean that they should. So while in some cases a network analysis can tell you a lot about your dataset, in other cases creating a network analysis and designating methodologies can skew the data and create relationships that are not there if the user is not careful.
Sources:
Weingart, Scott. “Demystifying Networks”. December 14, 2011. retrieved from http://www.scottbot.net/HIAL/?p=6279