The Perceptual and Cognitive Limits of Multivariate Data Visualization
In the article ‘The Perceptual and Cognitive Limits of Multivariate Data Visualization’, the author talks about the limits to the number of variables that a single graph can display depending on the approach that’s used.
The Visual perception and cognition of a graph imposes limits on the number of variables that we can encode in a single graph when we’re using a different visual attribute for each. There are a variety of factors that affect this:
- Only a few visual attributes work well for encoding data in graphs.
- Using some visual attributes eliminates the possibility of using certain other attributes in the same graph.
- Working memory can only attend to three or at most four chunks of information at a time, so limited value is added by including more than four
- Increasing the number of visual attributes in a single graph beyond a certain number creates a cluttered appearance that undermines perception.
Just to put that information into perspective, below is an example of a graph that I came across when I was doing research for a class in the past.
As it can be seen from the graph, there are way too many variables. It is confusing, and it is hard to keep track of what is going on in the graph. Going back to the article, the author stated that adding to many variables ‘creates a cluttered appearance that undermines perception’, which is exactly what is happening in the graph above. It is okay to add a variety of variables to your graphs, but it is important to ensure that the graph is easy to follow along as well as clear and specific.
