Check out a free preview of the full Building Custom Data Visualizations course:
The "Gestalt Laws of Grouping" Lesson is part of the full, Building Custom Data Visualizations course featured in this preview video. Here's what you'd learn in this lesson:

Shirley covers the Gestalt principles of proximity, similarity, and exclosure that help explain how to utilize how the brain processes patterns for designing data visualizations.

Get Unlimited Access Now

Transcript from the "Gestalt Laws of Grouping" Lesson

[00:00:00]
>> Shirley Wu: And once we've talked about marks and channels, another thing that I want to talk a little bit about is this thing called Gestalt Laws of Grouping. So apparently Gestalt is German for patterns. So I think something that came out, I think, in either the 60s or maybe earlier than that.

[00:00:20] That's kind of looking into the psychology of how the human mind just naturally groups individual elements into patterns. And there is, one of the books that'll I'll recommend later called The Functional Art by Alberto Cairo. He has a really great section talking about how to use these Gestalt Laws to our advantage.

[00:00:48] And in particular, how to use them in data visualization to help your readers save processing time. Because these are the things that our brains already naturally do. So utilizing them in the right context to help them read our visualizations easier. And so there's, I think, six of them.

[00:01:08] But I think we'll only talk about three of them today that I think are the most relevant to us as we design visualizations. And so the first of them are proximity. And so proximity, that's pretty straightforward. Proximity is that we tend to see things that are near to each other as being in the same group.

[00:01:35] And so Alberto's recommendation is just to put related objects near to each other.
>> Shirley Wu: To kind of emphasize that natural grouping. The second one is similarity. So the same looking objects will appear like they are in a group or we'll process them as being in a group. And so you can use this to indicate similar things.

[00:02:08] And this is especially helpful if those things can't be placed close to each other. Like, perhaps you've already used the proximity thing to highlight some other sort of natural grouping, that maybe are a little more high in importance. But then you'll also want to indicate this other sort of natural grouping, and maybe you will put that similarly.

[00:02:30] For example, these kind of bar charts are a great example. So it seems like there's already one sort of natural grouping here. Maybe these are, I don't know, years or something that's naturally grouped by it. And then the colors though, you know that those colors, the bars with the same colors probably belong to the same category of something.

[00:02:55] So, really, the similarity is really helpful in conjuntion with proximity to indicate these two different kinds of groups.
>> Shirley Wu: And the other example here is if you have geography. If you have a map, then proximity is not something that you can use, because you know, maps have position already encoded in it.

[00:03:16] So maybe using colors or textures or some other way to indicate the same thing is a great way to indicate groupings
>> Shirley Wu: And the last one I wanna say, is this thing called enclosure, which just means that our brains will tend to, if you box a certain part, or if you just give some shading.

[00:03:42] We'll think of that as somehow related to each other. Our brains would just go naturally towards those sections and think they're related to each other. So this is kind of helpful for creating visualizations that have maybe multiple sections that should be paid attention to. So these are the Gestalt Laws of Grouping.

[00:04:02] I like to think of them as if marks and channels are kind of the primary way to describe our data, then maybe these Gestalt Laws of Grouping can help us give the context around that data. Kind of give it that secondary layer of information that maybe can't be mapped to individual marks.

[00:04:22] Maybe they aren't to describe individual data points, but rather to describe groups of those data points together, those marks together. So that's the way that I like to think about these to kind of give. I think they are really great for helping give context.