The practical application of the Gestalt principles in data visualisation may seem obvious in many instances, but perhaps not all. Below are three examples of how Gestalt principles can be applied and our perception mobilised to increase the communicative value of visual representation.

What principles are applicable in each image?

Example 1: Genes

Ett färgdiagram som beskriver mönster i genuttryck spritt över flera prover

Creative Commons BY SA

The first third of "Gene set 1" (sample 1, 2 och 3) describes a higher gene expression than the remaining two-thirds of the same set.
Source: Thomas Shafee, Wikimedia Commons

The colour diagram above shows gene expressions spread across three different sets. Red indicates a higher level, white a median level, and blue a lower level.

Comment:

In this diagram, the principle of similarity has been employed at a very basic level to describe the orientation, shape and size of the different coloured fields (that is, horizontal rectangular fields of the same size). This makes it clear to the viewer that each field is, at least to some extent, equivalent. The same principle of similarity is then used to group as well as separate the fields through the use of colour. It is due to the grouping power of colour that the eye is able to quickly perceive the concentration of red fields in the upper left corner of the table, it is clear that the first three samples of gene set 1 displays a higher gene expression.

Example 2: Flowers

Cluster visualisation of Ronald Fisher's classic dataset about three different flower species, "The Fisher Iris Data" (1936).
Source: Alboukadel Kassambara, Datanovia.

Comment:

In this example, in a cluster visualisation of Fisher's classic 1936 dataset (above) about flower species, the proximity principle reveals two primary clusters, each clearly separated from the other. When we take additional variables into account, it turns out that one of these clusters is actually composed of two sub-clusters. By using color, the principle of similarity is employed and the three clusters emerge. The contour line takes advantage of the principle of enclosure and the three clusters appear even more distinct. When we furthermore allow the different species to be represented by shape (this is the principle of similarity, again), we can clearly see that the two secondary clusters actually consist of two iris species; versicolor and virginica.

Example 3: People

Age pyramid showing differences between men and women in Europe

Creative Commons BY SA

This population pyramid displays the distribution of European men and women by age groups. The data is from 2019.

Comment:

In this age structure diagram, or population pyramid, it is the lack of symmetry that enables us to easily perceive differences in population distribution. There are several principles at play here; similarity in the rectangular shapes as well as the use of colour, continuity in the vertical central line created by dividing the rectangles in two colours, common fate in the orientation of the rectangles. Finally, symmetry is both asserted and rejected; asserted by the visually strong central vertical line, rejected by the difference in length of the rectangles on either side of the central line.