[MUSIC] Hi, welcome back. Well designed visualizations need to work at many levels. Not only the visual elements, but also the scale indicators and text used for labeling and describing the data as well. As we have seen in a previous lesson, there are many ways that visual design elements can enhance or detract from the ability to correctly interpret the meaning in the data visualization. The same applies for language and scale decisions. For this final lesson, you will assess how your choices in language, labeling, and showing scale in your visualizations may affect your audience's interpretation of what is being shown. Let's take a look. Hi, and welcome back. While a picture may be worth 1,000 words, sometimes a few key descriptors can really help clarify and define what's going on in the picture. Or make the picture more confusing. Let's look at how language, labeling, and elements such as scale indicators can make or break the interpretation of a data visualization. Word choices and descriptors that are familiar, or at least have clearer meanings to your target audience, can help ensure that people understand the meaning and context of the visualization. Consider that your audiences may have very specific ways of wording things or understanding certain terminology that may not be universally shared. So, for example, a word such as risk might mean one thing to a person in the cyber security industry and something entirely different to someone in the financial services or insurance sectors. Without a basic awareness of your audience's potential industry standard vocabulary, and what they understand the words to mean for their particular context, you may be throwing them off by labeling alone. Beyond industry specific connotations, there can sometimes be substantial, geographical or cultural differences within even, say, the English language. Let's say you're designing a dashboard and want to show the new data in an area called New Today. The thing is that today maybe a clear and simple term but also can be tricky when you start to think about it. Many people might be looking at the display in different parts of the world. So someone's today might be another person's yesterday or tomorrow. On a more culture oriented level, if it's cold, you might want to put on a sweater in the US, but a jumper in the UK. It's the same kind of apparel, but a very different name. In addition to challenges with specific word meanings, there are other kinds of labeling problems that can be introduced into a visualization. For example, cryptic terminologies, unusual data labels, ambiguous data labels or acronyms that come from column headers in a source database can get introduced into your visualization. If so, these descriptors may not be clear to your end users or anyone other than the creators of the data table themselves. Along with word choice, considerations about scale selection can have a significant impact on users' perception of the data. For example, when representing a change in amounts, the numerical starting point of the change can have a big impact on the visual comparison. As in 1 to 4 versus 81 to 84. Percentage wise, the change from 1 to 4 is much greater than 81 to 84. So if they're put together and compared, it's helpful to see, visually, that change in percentages in the right framework. The uptick in both of these cases is three, but in the first case the percentage change is much larger. Consequently, the slope of the line showing the change should be steeper. Depending on the choices of scale of the visual impression of the comparative change, it can be either quite clear or entirely misleading. Although there are some advanced techniques for deriving comparable scales for different types of data, one simple rule of thumb is to avoid broken axes. Or, said another way, starting an axis not at a zero point to save space. Start your axes at zero when possible. Scale and comparisons can quickly lead to visual distortions, so keep it as simple and straightforward as possible always. Visualizations are a way to communicate meaningful patterns in data. While the visual component is generally the most obvious, part of the communication is also, often necessarily, words and numbers. If the words and numbers are not chosen carefully and thoughtfully, then the visual aspect of the design may not be understood. That's something to really think about. See you next time.
