What, How, and Why
For this sprint, we were given a data set compiled by the City of Seattle’s 911 Incident Response database (data.seattle.gov) that had multiple thousands of entries for a variety of events. I used this data to create visualizations that would help a hypothetical user (I chose a user in the police force) find an answer to a research question. My user wanted to be able to predict which calls might be false alarms. The purpose of my visualisations was to make it easier for the user to solve his/her problem and understand the data in a more intuitive way.
While creating the visualisation, my greatest difficulties were choosing the most suitable way of presenting and correlating the data, and using the software itself. The latter is easily fixed but my attempts at data correlation raised questions such as “what is the minimum number of dimensions and measures for a successful visualisation?”, “how many correlations should be shown (i.e. the map shows location and type. Should it also have included frequency?)”, and “how much information can we extrapolate from the data before we are simply making assumptions? (i.e. the Christmas example in the deliverable)”.
In the Future
We live in the information age, where more data than we can ever process is available at little to no cost on the internet. This, combined with a faster pace of living (and thus more scattered attention), makes data visualisation very important. For many people like me, it is much easier to understand a point and think about the data when confronted with a visual compared to a long spreadsheet. The technique could be applied in any situation where a large amount of information needs to be conveyed quickly. This may include climate data, real-time monitoring systems (such as at a power plant) and presentations. It probably would not be appropriate for small data samples or projects that are more dependent on qualitative observation (such as user research) as it would not serve its purpose of relaying a large amount of information quickly.