By @carnby (Eduardo Graells Garrido).
While the map visualization was aimed at continental analysis, this visualization is aimed at country specific analysis. M-Lab has different variables that together can give a clear picture on the state of internet quality on a country. This data has the form of time-series and it is available in different levels of detail.
This visualization uses the concept of small multiples created by Edward Tufte, which says that instead of putting different visualizations one on top of each other in a shared space, it is better to have small visualizations in a splitted space because this enhances comprehension and facilitates comparison of variables.
Following this principle, the visualization has two parts. The first one, the left column, presents the time-series at country level for each variable tracked by M-Lab. In this way, the exploration of trends in the data is natural and easy. Here we are not really interested in the particular value of a variable in time, but on how that variable changes.
The second part, the right column, focuses on the internet service providers (ISPs) for the selected country. Here, while this data is also in time-series form, I think it’s best to ease comparison between different providers using a different, localized approach. To do this, for each provider, a small visualization of bars is created, where each bar represents a different variable in a selected time frame (determinated from the interaction with the time series, which is not implemented). In this way, each group of bars encodes the state of an ISP in a certain period of time and thus it is possible to compare two different providers just by looking at this encoding. Exploration is encouraged because the time series on the left column are interactive, and when the user hovers the time series with the mouse pointer, the ISP visualizations react to reflect the state of the providers at that point of time.