Data Visualization

What is Data Visualization?

Data visualization is the process of representing data in visual form, which could include charts, maps, and graphs. Data visualization is a great way to extract insights from your data, as it is easier to identify outliers, patterns, and trends compared to other analysis methods — such as observing the raw data in a table. Visualization also allows us to further enhance our storytelling when working with data. Combining insights from data with visualization can help us build a narrative, bring data to life, and communicate a story to make complex, technical topics more accessible to a broader audience.

Benefits of Data Visualization

Three benefits of using data visualization are the ease of identifying insights, the ability to summarize large quantities of data, and the ability to communicate a clear message.

Identify insights

The human eye is naturally drawn to patterns and colors. We can quickly identify and distinguish colors from each other as well as see what is trending up or down. We can use data visualization to take advantage of this strength. Insights and messages that are hard to understand by simply looking at raw numbers can become easier to understand and interpret with the use of data visualization.

Outliers are more easily detected in data visualizations compared to rows and rows of raw numbers in a spreadsheet. A common strategy for identifying outliers or patterns is using summary statistics, such as mean and standard deviation. However, using these aggregations can hide patterns and there is a chance that we could miss important insights. Data visualization, with the use of images and color, make it much easier to detect this information. 

Summarize large quantities of data

We are faced with more and more data every day in both our personal and professional lives. Large quantities of data can be overwhelming. It can take time and cost us effort to try and derive meaning and information from this vast quantity of data. Data visualization can help us to summarize large quantities of data that come from multiple sources — like data warehouses, online sources, commercial systems, legacy Excel files, or anything in between. Data visualization allows us to compile all of this data and turn it into useful information. Different types of charts and visuals can help us to find and communicate insights that previously may have been lost in the volume.

Communicate a clear message

Working with data can be a complex and onerous task. Generating insights from data requires experience and skill. It can be a challenge to communicate these insights with others who do not understand the data and the limitations of our analysis. Data visualization provides a way to communicate a clear message from our insights, without us needing to explain (or more importantly, our audience to understand) the complex, technical methodology and concepts it took for us to find the insights in the first place.

Popular Data Visualization Types

When communicating messages from our data, there are a few popular types of insights:

• Comparing groups or categories with each other
• Exploring the relationship between different groups
• Following the trend of a value
• Highlighting the distribution and outliers in our data

We have popular data visualization types that can help us communicate these four key types of messages.

Bar charts

A bar chart is a chart that visualizes categorical data with rectangular bars. These bars can be plotted either horizontally or vertically. Bar charts are a great way to compare categorical data and are primarily used to see the count of each categorical variable. In this bar chart visual below, we can compare the mean adjusted close of three different stocks and quickly see the magnitude of differences between them.

Scatter plots

Scatter plots are charts that explore the relationship between two numeric variables. Scatter plots can allow us to see how one variable changes while another changes. This can lead us to draw conclusions about the correlation between two variables, or even if a causal relationship exists between the two. In the scatter plot below, we can explore the relationship between the percent change in stocks against the percent change in bonds over a two year period, to see if there is a linear relationship between the two.

Line charts

Line charts are most commonly used to display trends in a value over time. A common use case is in finance, to display stock price over time. This is the case in the line chart below, where we follow the adjusted close price of a stock index over a two year period. Notice how we can quickly identify when the price was the highest, lowest, the volatility of the stock, as well as the overall trend in price movement in the first half of the chart compared to the second.

Box plots

Box plots are a great way to show the distribution of our data and identify outliers that could cause us issues in our analysis. A box plot identifies the interquartile range and median of the data with the box and the relative maximum and minimum values with the whiskers. We can see the shape of the distribution of our data as well as any outlier values that fall outside of the whiskers.

Who Should Use Data Visualization?

Anyone that is working with data and wants to communicate their insights should be using data visualization to help tell their story.

Data analysts and financial analysts

Data visualization is a must to fulfill the day to day tasks of a data or financial analyst role. With data visualization, these analysts can easily identify key insights from their data to showcase to the company, see trends in their data to extract possible next steps the company can take, and build dashboards to easily display important information such as KPIs to stakeholders.

Data scientists

Data visualization is also a key component of a data scientist’s day to day tasks. With data visualization, data scientists conduct exploratory data analysis (EDA), which helps them extract insights from the data that will need to be used when creating machine learning models. Additionally, when troubleshooting models and improving model performance, data visualization will allow data scientists to understand their data more when creating models that are more suitable to the data.

…and everyone else

Technically, anyone despite their formal role should benefit from the use of data visualization. It is an intelligent way of displaying and analyzing data to make optimal decisions. When telling a story or in any decision making process, data visualization is a great way to convey facts and data to the audience that can support your point.

Additional Resources

Dashboards & Data Visualization Course

Dashboarding

The Business Intelligence Cycle

See all business intelligence resources