4 Easy rules to select the right chart for your data
Data contains knowledge that needs to be unveiled. This is why data becomes valuable only when we are able to detect the information it carries within.
But how do we do that?
How do we communicate effectively the data we own?
How do we provide the top management with a wider view and deeper understanding of the company status?
The answer to all these questions is easy: data visualization.
But what is not always so easy is choosing the right chart.
When making this selection we always have to consider the data set that we have, what we want to show, and the audience.
We know that we will use a different chart if we want to evaluate a distribution, make a comparison, study a composition, or assess a relationship. Even so, the options among the different graphs and charts we can use for each study is wide.
For example, when comparing value sets, we can use a range of graphs and charts, like: column, bar, circular area, line, scatter plot or bullet.
What are the features of each one and how do we choose between one or another?
We know that all of them are the optimal selection when we want to show a comparison between value sets. But how do we know which one to use among them all?
There is not one right option that will be the best one in every situation. Each chart can be the best response to a different answer. It always depends on what we want to show with our data visualization and what we are trying to achieve.
In order to choose the right one, there are two questions we need to answer:
- How many time periods are we considering?
- How many variables are we evaluating?
Assessing a Relationship.
When assessing a relationship between data sets, we are trying to understand how two or more data sets combine and interact with each other.
This relationship is called correlation and it can be positive or negative, meaning that the variables considered might be supportive or working against each other.
When considering two variables we use a Scatter Chart. This chart shows how much one variable is affected by another by using a Cartesian coordinate system. The existence of a correlation between the variables will be easy to identify when the data points come to make a straight line.
If the number of variables that we need to consider gets to three, then the best visualization to use is the Bubble chart. Very similar to a Scatterplot, the Bubble chart enables to show the third variable in the area of the circle of each plotted point.
Evaluating a distribution.
When evaluating a distribution what we want to find out is the existence (or absence) of patterns and their evolution over time.
If considering two variables, we use a Scatter Chart as described previously. If rather we are considering just one variable, the best visualization to use is the Histogram. This chart is very similar to a bar graph, but in this situation, the range of values is set into a series of intervals.
Making a comparison.
When analyzing our data we might be interested in comparing data sets to understand differences or similarities between data points or time periods.
Comparing different groups during the same time period can be done using a Simple Bar Chart or a Variable Width Bar Chart. The only difference between the two is that the second one allows comparing not only the height of each bar, but also their width.
If instead the time variable is important, and we want to compare the same group across different time periods, we would rather use a Circular Area Chart (also called Radar Chart). This chart is a smart way of comparing multiple quantitative variables as the number of axes equals the number of variables and they are arranged radially starting from the centre. Each variable value is plotted along its individual axis and all the variables are connected together to form a polygon.
Studying a composition.
When studying a composition we are interested to understand the smaller parts forming a group, and how these parts change or stay the same over time.
The most popular way to show composition is through Pie Charts as they draw attention to important information quickly. Even if this chart is very simple and eye-catching, we should not abuse it: Pie Charts are difficult to read when composed of more than 7 slices and therefore should be avoided.
When the time variable is important and we want to understand how a composition changes over time, Pie Charts are not the correct answer anymore. We would rather use a Stacked Bar Chart or a Stacked Area Chart. Both of them will enable the audience to understand how the composition of a data set changes over time by dividing the bars (or the area) among the categories in the data set.
iDashboards is a data visualization software that makes it easier for businesses worldwide to visualize the information they need and effectively communicate to their audience. Check out our complete guide on “top charts and graphs” and keep track of your KPIs through real time dashboards for a more accurate and efficient decision making. Click here to find out more.
