Data Visualization: Three Types of Memories

Humans are visual creatures. Research shows that we process visuals 60,000 times faster than text. Images are the easiest things to remember by our brains and are essential for our learning.

"90% of the transmitted information in the human brain is visual.While others animals, such as the dog is characterized by the sense of smell, or hearing sense in the case of bat; the human is undoubtedly a visual being"

There are three general types of memory; discussed below:

Sensory – stimuli are received through the five senses. Our brains process sensory input fairly quickly and automatically; this is called preattentive processing. Preattentive attributes are those that people can processes automatically such as color, position of scatterplot points, length of a bar, etc. For example, in a chart with gray bars, coloring something blue would help it stand out.

Short-term – the ability to remember and process information at the same time. This type of processing is called attentive processing because it requires you to actually pay attention and to remember something. An example of using short term memory is trying to remember your credit card number once you first see it, you won’t really remember it unless you make a conscious effort to commit it to memory.

Long-term – storage of information based on meaning and association (semantically). When trying to leverage a person’s long-term memory you can elicit an emotion or make it relevant to the audience. Typically, when designing visualizations, you should focus more on using sensory memory which would make it easier for your audience to understand. 

Cognitive load is the amount of effort that is required from short-term memory at a given time. Since this memory has limited space, the goal of data visualization is to minimize cognitive load

What are some of the ways to limit cognitive load in data visualization?

Provide a clear title/ subtitle that tells the audience what it is they are looking at

Remove any visual distractions such as borders, grid lines, zero lines, unneccesary legends, excessive encoding, etc

Don't use too many colors in the view (under 5 different colors)

Use consistent font type/ size

Use filters to allow the audience to view various sections of the data

Don't clutter the view with too many charts (3-4 charts per dashboard)

When building your next data visualization keep the three memories in mind!

References

https://ernestoolivares.com/2013/01/11/we-are-90-visuals-beings