Data visualization is a technique you can use to communicate data and insights in a way that makes it easy for your audience to understand, make decisions, and act. And that audience can even be you. Data visualization is a technique that can be used to glean as well as to communicate insights.
Due to its usefulness in both exploratory as well as explanatory data analysis, data visualization as a tool to communicate with data has gained increasing interest. Before getting started with data visualization designing, it’s important to be able to identify the elements of a chart and some basic key data visualization terms.
If you have worked with charts and graphs before, these may be familiar to you. This article addresses the nuts and bolts of data visualization starting with identifying the elements of a chart and defining key data visualization terms. The difference between exploratory and explanatory analysis is then discussed with an illustratative example pulled from healthcare patient experience publicly reported data.
ELEMENTS OF A TABLE
One of the fundamental components of data visualization is understanding the elements of a chart, including the title, axis, legend, value labels, and series
1 Y-Axis Label
The y-axis label is the vertical axis running down the left-hand side. It displays the values for one of the variables. In this example, the survey questions are displayed on the vertical axis.
2 Chart Title
3 Variable
4 X-Axis Label
5 Legend
TERMINOLOGY
Another important aspect of data literacy is understanding key terminology. The following terms are critical when it comes to the essentials of data visualization.
Term | ​Definition |
Data set | A collection of related data points |
Data Source | The origin of the data, such as a database or spreadsheet |
Axis | A line representing a quantitative scale for a variable |
Scale | The range of values represented on an axis |
Series | A set of data that represents a specific category or variable |
Value Labels | Numeric labels that indicate the value of a data point |
SEEING AS WELL AS SHARING INSIGHTS
Exploratory and explanatory visualizations are the two main reasons for creating a visualization. In the first case, you want to explore a new data set and uncover findings. In this case, you're creating the visualization for yourself. The second case involves presenting insights to others in a clear and compelling way.
In both cases you'll be using descriptive statistics. Descriptive statistics summarize or describe the characteristics of your dataset. They are only concerned with the data that you've already gathered, and the story being told by that data. A deeper dive into predictive or inferential statistics is typically relegated to data scientists.
As an example, here is data from CMS, the Centers for Medicare and Medicaid Services. This publicly reported data set provides responses from the 29 question HCAHPS survey measuring patient experience over a year.
I limited the data to hospitals in my state of California and looked at four questions. The four questions reflecting hospital environment, call light responsiveness, and getting help to the bathroom when needed. I selected these four questions specifically because they have been addressed in some way by varying volunteer programs in different hospitals.
Allowing Excel to create my chart, I got the following graph.
Excel doesn't know my business question, so it gives me a distribution of the responses represented as a colorful pie chart. The chart shows the distribution of the answers to those four questions. Each of the four questions could be answered in one of three ways. The answer was either "always", " usually", or "'sometimes' and 'never'".
EXPLAIN OR EXPLORE USING DATA VISUALIZATION
That results in a total of 12 responses which get lost in the pie chart. Insight is very difficult to glean.
However, by converting it to a horizontal bar chart we start to be able to see some insights. A vertical bar chart is an excellent tool because the percentage’s sum to 100%. Horizontal bar charts are a great alternative to pie charts for this reason.
The visualization improvement from the first chart to the second provides much more valuable insight but it's not quite ready yet for data sharing. By further reducing the clutter and intentionally adding back key design elements our result might end up looking something like this.
Data visualization is a useful tool to both explain or explore data. An example using publicly reported HCAHPS patient experience data was used to illustrate this simple, yet profound difference. By understanding the basics of data visualization and key terminology, professionals can make more informed decisions and communicate effectively with their colleagues and stakeholders.
Roseanna Galindo, ECBA, CAVS
Roseanna Galindo is Principal at Periscope Business Process Analysis and a champion for data literacy and the human experience in healthcare. Learn more about Roseanna and her blog, The Periscope Insighter, by reading the opening post, Venn The Time Is Right.