This week I visualized the most popular baby names in New Zealand from 2001–2010 using an bumpchart that tracks how the popularity of different names rises and falls over time. In creating this visualization, my goal was to represent temporal continuity and relational shifts in a way that helps reveal patterns that might otherwise be difficult to see. This process reflects a point from Data by Design, which notes that while many say “the purpose of data visualization is insight,” true insight often emerges through a “process that is far longer and more hard-won.” My visualization attempts to identify how names move through years as part of a connected story rather than isolated statistics.
One decision I made to improve clarity was to reduce visual clutter by limiting the number of overlapping flows and adjusting colors so that major trends over time stand out. The alluvial form felt appropriate because it shows continuity and change simultaneously, which is difficult to achieve with simpler bar charts or static tables.
The process of designing this graph also made me reflect on the limits and responsibilities of visualization as a form of knowledge. In Data Feminism, one point from the third chapter is that data and the categories we choose to represent them shape what counts as evidence. The authors write that without the right categories, the right data cannot be collected, and without the right data, there can be no social change.
At the same time, this visualization also made clear the limitations of the dataset and the chosen form. Because the dataset includes a large number of variables, multiple names across multiple years. The resulting output becomes visually dense and difficult to read. Even with adjustments to color and layout, the bumpchart can quickly feel overwhelming rather than clarifying. Without further data wrangling, such as filtering to fewer names, aggregating categories, or restructuring the data to reduce crossings, there is only limited room for improvement at the visualization stage itself. This experience reinforced the idea that clearer visualizations often depend less on styling choices and more on careful decisions made earlier in the data preparation process.