The speaker:
Michael Neutze has a degree in Geography and has been with Destatis for 21 years, with experience in publishing, data visualisation, cartography and the census. He is known for animated population pyramids, traffic accidents calendar and interactive maps using open web standards.
The session:
People are fond of maps and maps are a statistical diagram that excites more than a pie chart. However, we need to understand what we do when we map data. How does a map become meaningful is therefore the topic and entry into the mapping topic. This also includes when not to make a map.
Secondly, mapping is a craft, it consists of using a map projection (since the earth isn’t flat) finding a suitable map-type for the data at hand and care for all the details that lead to correct interpretation, like colour- and size-keys, objective classification and an overall visual hierarchy for symbolisation of the underlying geography (e.g. borders, rivers, motorways) versus the statistical data.
Statistical organisations most often use choropleth maps in their day-to-day work: Usually a single variable, that represents relative values (density, concentration), is classified and administrative areas are then coded in shades of colour accordingly. It is this classification (quantiles, equal distance, Jenks natural breaks, ck-mean) that needs to be understood to make a meaningful map.
On the other hand, absolute values are mapped by using proportional symbols where the area of a symbol is proportional to the absolute value of the data. However, these symbols may overlap in a way that makes interpretation difficult. A force layout comes to the rescue, where simulated forces spread out the symbols but at the cost of losing location accuracy. We will see how these trade-offs can be dealt with.
Finally, with publishing online or even online first, whole new avenues into mapping open up. With a few examples from recent publications of Destatis, the German Statistical Office, we will see what kinds of interactivity are possible and where exactly interactivity and animation serve a better understanding of the data.