The Best Types of Graphs for Data Visualization

Data visualization

Imagine that you would like to convey relevant information to a group of people. Do you agree that depending on your audience, your narrative may change? Colors, texts, and even the presentation itself may vary according to the interlocutor. A more analytical profile requires greater granularity in the details, while a more playful audience may be more impacted by visual elements. In our context, the main means used to express data and make presentations are graphs.

What is a graph?

A graph is a means of representing something that can be measured. Imagine a graph that shows your twelve salaries for the year, with the months and values received. Increases and discounts can be seen. It is a very simple way to exemplify how this resource can help in the analysis of a context. There are simpler graphs, such as bar and line graphs, and some that are a little more complex, such as heat maps and scatterplots. Next, we will talk a little about the main types of graphs and how they can be used.

Types of graphs

Bar Graph

Let’s start with the bar graph. It is one of the most popular and is commonly used in analyses. One of its advantages is its objectivity. At a quick glance, it is possible to perceive which information is being displayed and how to interpret it. Be careful when one of the axes (X or Y) has many elements. This can make visualization difficult. Bar graphs are indicated when it is desired to easily compare the values of the bars.

Line Graph

It is usually used to measure a value over a period of time. It is recommended to highlight trends in a given period, that is, whether the displayed value is increasing (upward line) or decreasing (downward line). In the example below, we see monthly sales over the months of the year and can easily identify when there was an increase or decrease in sales from one month to another.

 

Heat Map

The heat map shows, through the intensity and variation of colors, trends of a variable. In the example below, the higher the value, the closer to green, and the lower, the closer to red. Analyzing the heat map, we see that November 2016 has the highest recorded value (strongest green). On the other hand, July 2017 has the lowest index (strongest red). Without the aid of colors, it would also be possible to obtain this information, but they make the analysis much more intuitive and quick.

 

Table

Another common graph is the table. It is a very objective way of representing data, and its advantage is that it allows the inclusion of a slightly larger number of variables for analysis. But be careful not to overdo it to avoid making the visualization experience confusing.

Scatter Plot

The scatter plot is used to identify the correlation between two variables (crossing of values on the X and Y axes). The further to the right and up the point is, the higher its value in both variables. It can be used to analyze, for example, sales effectiveness or the performance of a solution adopted by the company.

KPIs

In addition to the graphs mentioned, key performance indicators (KPIs) also deserve mention in our narrative. Through them, it is possible to highlight one or more values numerically. They are excellent for showing high-level numbers, such as sales made in a given period. Below we can see an example of the use of KPIs combined with line graphs to create a visualization of four segments. For each of them, purchase and sales values are shown through KPIs, and the purchase trend is represented through a simplified line graph. Thus, in a small space, it is possible to effectively and organizedly represent a series of data.

Using Graphs Effectively

Graphs are fundamental visual tools for good data analysis. Here are some tips on how to maximize the use of graphs for more effective analyses:

• Use a consistent color palette so that your data can converse with each other;
• Align the graphs and distribute them harmoniously on the page space, as this makes visualization more pleasant;
• Keep total values in focus to make analysis more intuitive (the top of the page is a good place for this).

The image below presents some of the mentioned suggestions.

Note that there are four pie charts in the example above. Here we have a good use of these graphs, as they show only a percentage in relation to the whole. However, this type of graph (as well as the pizza chart) should be used with caution. Many represented values can make the visual confusing and difficult to read, as in the image below.

Conclusion

We have discussed some ways to display data through different types of graphs. Although there are best practices, it is not possible to determine precise rules. It is recommended to know the purpose of the information well in order to establish which graphs to use and their details. Know your user well and make sketches before developing your final version. Feedback is very important in this process and will be essential for your visualizations to have good results.