What Data Visualization Is and How to Get Started In This Field

From COVID-19 data dashboards and election results to personal finance apps and social media analytics, visual representations of data are increasingly ubiquitous in day-to-day life and in business. If there are numbers involved, it’s often expected that those numbers will be visualized in a chart or a graph.

(Getty Images)

“Ten years ago, we definitely saw people who specialized in business intelligence or data or analytics, and that was their job, and everyone else was expected to send a request to those people and wait for reports back,” says Ellie Fields, senior vice president of product development at Tableau.

That’s no longer the only option. With a plethora of online tools and software packages available, and online classes and resources to help you hone your skills, the barrier to entry for creating compelling data visualization is lower than ever before. That means you can become proficient in presenting data visually without a background in computer science, data science or programming.

But with so many options, finding a place to start can be overwhelming. In this guide, we’ll give you the rundown of how you can start creating charts, graphs, maps and reports for your business or organization – or if you’re thinking about pursuing data visualization as a career or personal hobby.

Businesses large and small use data every day to make decisions about their inventory, sales and investments. From sales reports to human resources data, hard numbers help businesses track growth and analyze trends that help them make strategic decisions for the future. Business analytics includes everything from data aggregation (the process of gathering data) to forecasting and predictive analytics, which provide businesses with insight on what could come down the road. According to a 2018 MicroStrategy study cited by Harvard Business School, companies have boosted process and cost efficiency by 60% using data.

Business owners and executives may have a “mental story” about how their business works, their costs and where they’re succeeding, Fields says, but data often tell a different story. “When you can incorporate data into the way that you understand your business and the decisions that you make about it, you can be much more effective,” Fields says.

And there’s more data being captured today than ever before, with the widespread use of smartphones and other technology. An estimated 41 zettabytes (each zettabyte a trillion gigabytes) of data were created, captured, copied or consumed across the globe in 2019, and that total is expected to continue increasing, reaching 59 zettabytes in 2020 and 149 zettabytes by 2024, according to Statista.

But it’s often hard for numbers and data, even in more comprehensible smaller chunks, to tell a story in their raw form on a table or in a spreadsheet. “People think in stories, and using data as part of that story is really important,” Fields says. “For most people, if you literally have an Excel sheet of even just 20 by 20 rows and columns, it’s hard for the human brain to parse. Our brains just don’t work that way. And even if you’re really smart and good with numbers, it’s hard to get an accurate picture of what’s going on there – to find trends and outliers.”

Visualizing data – whether in charts, graphs or some other form – is important because it can give data meaning to a broader audience. “Visualization gives us a way to parse and understand data so that we can add it to our stories, we can incorporate it into our thinking,” Fields says.

Growth in big data and the increasing importance of data to business strategy makes it a growing field – and one where labor is in short supply. As demand for data increases, jobs analyzing and visualizing data will, too. LinkedIn lists data analysis, analytical reasoning and business analytics as some of the top skills companies are searching for to help their organizations make data-driven decisions.

Data visualization – like the underlying data itself – comes in many shapes and forms, but you’re probably already familiar with some of the most commonly used types of visualizations, even if you’re not familiar with their names.

The Most Common Types of Data Visualizations: Charts and Graphs

Below are some commonly used types of data visualizations. All of them are charts, which use symbols including icons, lines, bars, points, areas, colors and connections to bring meaning to data.

  • What it is: A line chart or line graph uses points connected by lines to display data and comes in a number of styles and variations.
  • When it’s used: Line charts are most often used to display data over time.

  • What it is: A pie chart divides a circle into pieces resembling slices of pie to show proportion, with the size of each slice showing the amount or quantity it represents.
  • When it’s used: Pie charts are most commonly used to represent percentages of a whole.

  • What it is: In a bar chart, two or more rectangles are used to display variables, with the height or length of the bars corresponding to the value of the variable.
  • When it’s used: Bar charts display and compare amounts of different variables. They can also be used to display data and trends over time when the bars are placed in order along an axis representing time.

  • What it is: A scatter plot displays two variables of the same data point along an x and y axis. For example, a point on a scatter plot could be positioned on the x axis for an individual’s height and the y axis for the same individual’s weight.
  • When it’s used: Scatter plots are used to try to identify trends or the relationship between two variables.

Variations and Other Types of Data Visualizations

There are so many other types of data visualizations and variations that it would be difficult to list them all, but here are some other examples:

  • Histogram. A histogram is a type of vertical bar chart that shows the frequency of distribution. The bars correspond to a range of values and the height shows the number of data points that fall within that range. Unlike a bar chart, in which the bars can be rearranged, a histogram cannot be rearranged.
  • Area chart. An area chart or area graph is similar to a line chart, except the area between the line and the axis is filled in. Like a line chart, area charts are great for showing trends over time, and the addition of the filled-in area emphasizes differences between the data points.

  • Maps. Maps can be used to display data. Geographic areas can be shaded with the opacity or color of the shading corresponding to a range or value. Data can often be displayed in different ways. For example, here’s a chart that displays the same data both on a graph and with a map.
  • Treemap. A treemap uses shapes, usually rectangles, with the size corresponding to a value and each rectangle representing a data point. The rectangles can be nested to create categories and subcategories.
  • Infographic. Infographics cover pretty much any visual representation of information. This might include multiple graphs, charts or other information.

Here are some other types of visualizations:

  • Cartogram
  • Heat Map
  • Matrix
  • Steamgraph
  • Box plot
  • Dot plot
  • Sankey diagram
  • Radar chart
  • Chord diagram
  • Network graph

There are several data visualization tools available, from simple to complex. Some are free, some are open source and some require paid licenses. Different tools serve as industry standards, including Tableau, Microsoft’s Power BI, Domo and QGIS, along with many others.

  • Tableau is a widely-used visualization tool. It’s a software package that can be used for free (if you’re willing to save your visualizations to Tableau Public) or with a license. It can connect with Excel files, text and larger scale databases. For people new to data visualization, Tableau offers powerful capabilities with a drag-and-drop interface and without coding. While it’s pretty simple to use Tableau to create basic visualizations, it can be more difficult and has a larger learning curve when you need more advanced visualizations. It can be used for pretty much anything, from business analytics to journalism or for making data visualizations as a hobby. There’s a large community of data viz creators who use Tableau, so there’s plenty of support online.

  • Microsoft’s Power BI is oriented toward businesses for business analytics, but it’s powerful for analyzing business data. Users can connect to various data sources, and it allows for customizable dashboards and detailed data reports. It’s also integrated into Microsoft’s product suite and works well with Excel and Azure. It is a paid service, but it’s more economical than Tableau’s paid licenses.

  • Flourish is another excellent data visualization tool used by businesses and journalists, among others. It’s entirely online with no need to download software, and it has a wide range of templates and examples. For people new to visualizing data, Flourish is easy to use, has a ton of customization and interactive capabilities, and offers a free version. There’s also a special free license for journalists and newsrooms. For business use, you’ll have to buy the paid version.

  • Microsoft Excel’s built-in chart function
  • ChartBlocks
  • Datawrapper
  • Adobe Illustrator
  • ArcGIS
  • Mapbox
  • Charticulator
  • Infogram

You can explore more tools at vistools.net, where you can sort tools based on your programming skill, operating system, features and pricing.

With so many data visualization tools on the market, learning a programming language or coding is no longer a requirement. “There’s a huge amount of really interesting and thoughtful viz done without necessarily programming,” says Tamara Munzer, an information visualization expert and professor of computer science at the University of British Columbia.

The question, Munzer says, is whether you need something custom or bespoke, or to take existing charts and templates and put them together thoughtfully and carefully to tell the story you mean for them to tell. If it’s the former, it may be helpful to learn a programming language.

But in most cases, the tools that are already on the market can help you accomplish your goals without coding. “Thirty years ago, yeah, I had to program,” Munzer says. “But we are not 30 years ago, and there are a number of tools that are trying to democratize visualization.”

R and Data Visualization

Whether or not you’re visualizing data, R is a free, open-source and widely used programming language and software environment for statistics and data. It’s great for cleaning, mining and analyzing data. There’s an enormous online community using R, so there’s plenty of help and inspiration. One of the best parts of R is that it’s highly extensible, meaning that there are add-ons, called packages, which add to its capabilities. Two of those packages are ggplot2, used to create static visualizations, and Shiny, which can be used to create interactive visualizations and dashboards.

Python and Data Visualization

Python is a programming language with a variety of uses well beyond data visualization. It’s often used to gather, process and analyze data. It’s flexible and relatively easy to learn. Like R, Python has add-ons, called libraries, that are used to visualize data. One that’s often used is Altair, which allows for a wide range of visualizations.

Here’s an example of what this language looks like using the Altair library.

import altair as alt
import pandas as pd
source = pd.DataFrame({
    'a': ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I'],
    'b': [28, 55, 43, 91, 81, 53, 19, 87, 52]

The previous code produces this result:

Javascript and D3

If you’re particularly interested in data visualization on the web, you might consider learning JavaScript, a common programming language used on web pages. D3, a powerful JavaScript library, is often used for creating interactive web-based visualizations. D3 has a high learning curve and requires a pretty good understanding of the JavaScript language with the upside of being highly flexible.

If you’re creative, like numbers and have an interest in identifying trends and insights from data, you might want to learn more about data visualization. While you may not need to learn a programming language, you’ll need to be willing to learn by trial and error, and you’ll have to decide what your goals are.

But first, you need to be able to understand data and how to present it without misinterpreting that data and misleading those you’re trying to reach. For example, you’ll want to know the difference between a median and an average and how average can be misleading if there are outliers in your data. If you don’t have a good grasp of data itself, your data visualizations may be incomplete or inaccurate. But that doesn’t mean you should be afraid to begin working with visualizations.

“You know, there’s sort of a delicate line between cautioning people appropriately and making them afraid to take any step forward for fear that the rattlesnakes will bite,” says Munzer, who wrote a book about working with data visualizations. She also recommends “Data Feminism,” a book about data ethics inspired by intersectional feminism. There are plenty of online data literacy courses and resources to get a basic understanding of data, including this free course from Tableau.

There’s also the option of pursuing undergraduate degrees in data science or business analytics, though many professionals who work with data visualization every day don’t have a degree in those fields. There is still room for degrees, especially for people who want to dive deep into data as their career.

“The vast majority of people working in data today, and making use of data, do not have computer science or data-oriented degrees. They have degrees in everything, and are doing jobs doing anything,” Fields says.

Without pursuing a formal degree, you’ll need to set aside time to learn on your own, but there’s a range of courses, books, tutorials and other online resources for getting started in data visualization. Here are a few:

Source link

Leave a Reply