Data Visualization on a Budget: Simple Tips, Techniques, and Low-Cost Tools
by Linda Hofschire
The saying “a picture is worth a thousand words” certainly rings true when it comes to data visualizations. Recently, we have been bombarded with statistics about COVID-19, and some of the most impactful and digestible information has come in the form of visualizations—for example, the ubiquitous “flatten the curve” charts.
|In addition to maximizing the impact of your reporting, you may also find that visualizing data can be fun.
As library and information science professionals, it is unlikely that we would find ourselves in the position to be reporting public health statistics. However, many of us regularly report data about our work, including operational statistics (circulation, database use, etc.) and the results of user research, to demonstrate the impacts of our services. How can we take advantage of the power of data visualization to maximize the impact of this reporting?
The term “data visualization” may bring to mind images of complex infographics or interactive visualizations. However, for the purposes of this article, I’m using this term to describe a variety of visual presentations of data, including simpler techniques such as creating a chart in Excel or Google Sheets and formatting a number to make it stand out from a sea of text in a report. While most of us don’t plan to design complex visualizations, many of us share numbers in formats including reports, presentations, and social media posts. We can easily improve this reporting and make it more impactful by applying a few simple data visualization tips and techniques.
Defining Goals and Objectives
When designing a visualization, start by determining your main goals; these will guide all of the subsequent decisions you will make during the design process. What do you want your audience to learn or do as a result of viewing the visualization? As an example, let’s say that your library provides STEM programming for teens and needs funding to continue this effort. This could be the goal of your visualization—to demonstrate the effectiveness of the program in order to attract funding.
Next, identify your primary audience for the visualization. Continuing with the STEM example and based on the goal that was specified, your audience is potential funders for the program. Perhaps there is a local foundation that has identified workforce development for teens as a priority area. This organization could be your focus as you design your visualization.
After that, consider what data you have about the topic of your visualization. For the STEM example, you may have data about resources invested into the program in terms of money and staffers, data on program attendance, survey results about the impact of the program, observations from staff members about the program’s effectiveness, etc.
A key part of this step—after determining what data you have that are related to your visualization—is to narrow down your data. You should avoid presenting all of the data you collect, as this will overwhelm your audience. Instead, present the key datapoints that best support your goals and are most relevant to your audience. For the STEM example, this includes any data that address the audience’s priorities, such as survey results indicating that teens learn about STEM career pathways as a result of participating in the program.
After selecting your data, your next step is to determine how you will present them. Options include text, graphics, and charts—or a combination of these elements. Think about where you can take the opportunity to show—i.e., visualize the information—rather than just telling. It’s tempting to want to explain your numbers with a lot of text, but you should take advantage of the power of visuals to convey your message.
Text and Graphics
If you plan to use text or graphics to present your data, consider whether you can provide any context to make them more understandable to your audience. Here are some tips for adding context.
Per capita —While large numbers may seem impressive, they can also be abstract. To provide more context, you can present them in per capita terms (per person). For example, my organization (Colorado State Library) has a publication called Quotable Facts About Colorado Libraries. In it, we use per capita numbers to present annual circulation figures for the state: “123 million items circulated annually, or more than 22 items for each person in the state” (see Figure 1).
Ratios/fractions —Ratios and fractions can serve as a nice alternative to percentages to provide more context for a statistic. For the STEM example, if you had survey results indicating that 80% of participants learned something new about STEM career pathways, you could present this as “4 in 5 participants learned something new. …” This technique works well when using graphics to visualize your numbers, as you can see in our quotable fact about the number of Coloradans with library cards (see Figure 2).
References to pop culture —Can you compare your data to a pop culture phenomenon that many people would be familiar with? One of our quotable facts is that there are six times as many libraries in Colorado as Starbucks stores (see Figure 3).
If you create charts to present your data, it’s important to choose the chart type that most accurately represents them.
- A pie chart can be a good choice if you are providing percentages for four or fewer categories—for example, library cardholder status (yes/no) and library type (academic, public, school, other). The reason that you don’t want to go much higher than four categories with pie charts is that they become difficult to interpret if the pie is divided into numerous segments. You should also use a pie chart only if your categories sum to 100%, as the purpose of a pie chart is to present parts in relation to a whole.
- If you are presenting data with more than four categories, or if your categories sum to more than 100% (this can happen if respondents can “select all that apply” on a survey question with multiple response options), bar charts are a good option. Bar charts are versatile and can be used for other purposes, such as to demonstrate comparisons over time.
- Line charts are meant to show trends, so you should only use these if you are presenting data over multiple points in time.
After you’ve created your chart, there are a few design tweaks you can apply to make it easier to interpret:
- Add data labels so that the viewer doesn’t have to estimate the value of each chart element (bar, pie segment, etc.).
- To maximize white space, remove all unnecessary information, such as the axes and gridlines (lines that extend from the horizontal and vertical axes), and round data to whole numbers.
- Make your chart 2D, as three dimensions distort the proportions of the chart and make it difficult to interpret.
- Reposition legends so they’re as close to the datapoints they’re representing as possible.
- Use your title strategically. Many chart titles simply provide descriptive information (for example, STEM Program Survey Results). For more impact, make your title a sentence that captures your main point. By explicitly stating it, you won’t need to wonder whether viewers will reach that conclusion on their own.
See Figure 4 for a comparison of how applying these techniques will work to your advantage.
Graphic Design Tips
Now that you have determined how you will present your data, consider how you can apply some simple graphic design best practices for white space, color, and fonts to add visual appeal and impact. It is important to maintain some white space, which is the portion of a visualization that doesn’t contain any elements (text, graphics, etc.). This provides a visual break for the viewer and helps to emphasize your key points.
If your organization has branding guidelines, these should be your starting place for choosing colors and fonts for your visualization. Limit your color choices to three or four, and be purposeful in your use of them. You can use one or two colors as the main colors and one or two as accents. Another option is to choose one color for your visualization and use multiple shades of it. This technique helps your visualization appear clean and cohesive, and it eliminates the need to select multiple colors that coordinate well with each other.
For fonts, limit yourself to a maximum of two. Using a large number of them is disorienting and appears disorganized to your audience. You can use one font (and/or a larger type size, different color, or bold text) for titles and subtitles and a second font for text. This helps the viewer easily navigate the visualization, as the different fonts signify different content types.
You can also use color and fonts strategically to emphasize the main points of your visualization. For example, you can use one of your main colors to highlight certain graphics, text, or a specific part of a chart and then use a lighter shade of that color or gray for surrounding elements so that the highlighted information stands out (see Figure 5).
Similarly, you can emphasize text by making it bold, using a larger type size, or including a different color than what is used for other text.
Data Visualization Design Tools
At this point, you have specified your primary goals for your visualization and identified your audience, as well as selected your data and determined how you will present them. Now it’s time to choose a tool to create your visualization.
There are a variety of free or low-cost options for designing visualizations. Here are a few.
Consider Microsoft Office 365 or Google Sheets/Slides. These products are useful for creating charts as well as for designing presentations, reports, and infographics. Microsoft Office 365 has an icon library that contains many common subjects, including technology, communication, and education. You can customize the icons for your visualization by resizing them and applying your color scheme.
Another option to access a wide range of icons is the Noun Project website. It features an app that can be used with a variety of software, including Microsoft Office 365 and Google Docs/Slides to seamlessly customize and add any icon from its vast library to your visualization.
While Microsoft Office and Google software provide many options for designing visualizations, one potential downside is that you are starting with a blank canvas. If you don’t want to build a visualization from scratch, here are a couple of options that offer customizable templates.
Infogram is an online tool, available freely as well as on a subscription basis with more features, that contains templates for infographics, reports, slides, social media posts, and more. Infogram also offers a variety of chart types and icons that you can add to templates or use to build a visualization from scratch.
Adobe Spark is a freely available online tool that allows you to create a variety of visualizations (infographics, reports, social media posts, etc.) and video slideshows. Adobe Spark contains a library of stock images, icons, and music that you can select from to customize your visualization.
After you design your visualization, review it critically with an eye toward any opportunities to simplify or remove information. Remember, your goal is to communicate a point—not to include all of the data that you can fit into the visualization. One way to simplify your visualization is to break it into multiple parts. So, instead of presenting all of your information on one page or in one slide or social media image, consider whether you can present one point per page, slide, or image to make the information more digestible.
The next time you’re reporting data, think about how you can take advantage of the power of visualization by incorporating some of the tips discussed here. In addition to maximizing the impact of your reporting, you may also find that visualizing data can be fun.