The lecture from Lin and activities that we did based on data visualisation has been one of my favorite topics that we have gone through so far. One of the arguments that is relevant in the not only the digital humanities world but virtually any world that uses data, there are so many different factors that can go into how data is perceived and understood. Individually, I feel pretty comfortable with data in the sense of what to look out for and how to interpret things in a manner that is (to a certain extent) objective. However, there are so many little tweaks and nuances that can make data look like something that it fully isn’t, or more often show things that may have been present in the data but at an increased extent. Lin also emphasized how the colors, sizes, and shapes on a graph can have a large impact on how graphs are understood. Particularly, it was interesting to learn about how using different sizes is not usually a super effective way to show contrasts.
On working on this assignment and continuing to think about data visualisation, I really enjoyed looking at examples of what makes an example of bad data visualisation.
The graph above is pretty busy and difficult to understand. There is no real Y axis and the pictures that are supposed to represent different professions make it feel busy. On top of that, the scale of the X axis is misaligned in order to better show the trend that they are going for. At the start of the graph, there are eight years between the points, but by the end there is only one year between the points. Tiny switches like these can make all of the difference in making a graph show what you want it to show versus what it actually shows. I think that learning how to critically look at different data visualization will only become increasingly important in an increasingly digital world.
3 thoughts on “Blog Post #4: Data Visualisation”
Comments are closed.
Evelyn you raise a great point about the chart being too busy. When I look at it, I’m too distracted by the doctors looking down and the silhouettes of people looking up to notice the horrendous scaling. Additionally, I can’t stop thinking about the intervals for years along the X-axis. It’s so misleading. I think it’s good though we are talking about this as a class and can now be more aware of misleading graphs.
I like your comments and arguments about the graph from Washington Post being too difficult to understand! Personally, I was confused by whether the bars were on-scale vertically or not. The fourth last bar said 50k and the third last bar said 54k, but the gap between them seems much smaller than the gap between the third and the second last bar (58k), not sure whether it’s visually misleading with the various figures or because of the graph maker did not pay attention. In addition, I disliked the figure at the top of each bar – I spent a minute or two staring at them trying to figure out what it represents, and I also doubt why they are there since every bar seems to describe the same object.
You make a really interesting point of focusing on what not to do to learn what to do. The graph is definitely too busy especially with the lack of and confusing axis labels. I feel that a lot of what we learned in Lin’s lecture could be applied to this data visualization. For the story that they are trying to tell with the income differences in occupations, adding those labels and some sort of contrast (like color) should help shape the graphic. As you mention, I think it’s important to learn from the mistakes we don’t like about other ones so that we are able to better practice for what we should do.