[MUSIC] Hello and welcome to this module on Design for Understanding. Making sense of large multidimensional data sets can be a challenge for anyone. Your task as a designer is to make good decisions about encoding, arranging, and presenting data to reveal meaningful patterns and stories for your audiences. Throughout the other modules, you've already seen how getting the most from data visualization begins with keeping a few features of the human brain in mind. Recall that System 1 involves automatic and relatively immediate perception. While System 2 is a slower and more deliberative thought process. As a data visualization designer, you need to work with both of these systems in a way that enhances people's ability to both clearly perceive and accurately think about the data. But the process doesn't end there. For this module, let's take these steps a bit further and consider visualization strategies tailored for target audiences and the types of data that they are using. As you create your visualizations, think of yourself as a kind of architect. Your basic building materials include visual and coding elements such as color, shape, and size. Your task is to assemble these elements into forms that effectively leverage people's perceptual abilities. But the story doesn't end there. Just as an architect needs to consider factors ranging from construction materials and location, to people and the purposes for which the building is intended. You need to consider your materials, contexts, audiences, and purposes for your visualizations. Your designs may begin with visual perception principles, but need to finish by enabling a clear and accurate understanding of the data by your audience. In this lesson, let's explore putting together some building blocks. Selecting good visual encoding options in combinations is a bit of an art and a science. To illustrate this point, let's start with a simple example using encodings for color and shape. In this case, we can see there are two primary groups. One orange and the other blue. Within the two groups there are two subtypes based on shapes. In this example, circle and squares. You may find that the colors pop out at you more immediately then the shapes. But the two kinds of encodings don't interfere with each other. How much more can be added to express greater depth of information in detail without overloading the user? As we will explore, the answer to that will depend on many factors. Shapes, colors, and other design elements are kind of visual alphabet that can be mixed and matched to express compact, but rich and detailed visual stories. Say with additional colors, shapes, or lines. Different kinds of visual encoding options have different strengths and weaknesses. For example, as we have seen, colors are a powerful way to make patterns stand out. And to distinguish different categories or groups. However, it is not the right choice for presenting certain kinds of precise quantitative measures. For that kind of precision, attributes like length and 2D position are needed. That said, color can be helpful in giving qualitative sense of certain kinds of data. One common approach is showing a level of intensity or activity with a darker more saturated color indicating higher activity. And a lighter less saturated color showing lower or less activity. Even one of the simplest elements, the line, can be modified to communicate different quantitative and qualitative measures. Let's take a pair of lines and modify them to see what we can do to make them express different aspects of the data. My length can be used to show precise quantitive differences. Whereas line thickness can be used to show qualitative distinctions in the relative strength of connection between two entities in say, a network graph. Line color differentiates categories while 2D position can provide a quantitative representation of, for example, change over time. How many visual attributes can you think of for showing either quantitative or qualitative measures? Here is a list of a few common elements. Length, position, 2D position, orientation, line width, size, shape, hue, intensity, and others. We've only scratched the surface of this topic. And to learn more, I'd recommend reading discussions in books by Colin Ware, Stephen Few, and Jacques Bertin. Now that we've seen and reviewed some of the basic encoding elements and combinations for data visualization, let's begin to think about how to put them together for target audiences. You should have a sense of who, when, and where of your visualizations. And what kind of questions will be helpful for you to ask and be able to answer in the process of creating your designs. See you next time.
