Data Visualization: Choosing the Right Graph/Chart Type

Aug 20, 2024

In today’s Data Storytelling Visualization journey, we learn to avoid making the same mistakes as the past; not every graph/chart needs to highlight groundbreaking insights, and we must deal with the second-hand cringes.

Data graphical representations are an effective method of communication to the audience for any data-driven organization. However, the diversity of styles present makes it challenging to decide which one to use for a given set of data. Before getting into the content, let’s first distinguish between graphs and charts, which are often used interchangeably but are different visuals.

A chart is a visual representation of data in the form of graphs, tables, or diagrams / pictures, while graphs represent mathematical relationships between information sets, such as a grid with x, y, or z axis. In other words, all graphs are charts, but not all charts are graphs.

Step 1: Determine the Type of Data

Before selecting a chart type, you must determine the data type, which can be divided into four categories: quantitative, categorical, temporal, and spatial.

Quantitative data refers to numerical values, such as sales or inventory figures, while categorical data refers to non-numerical values, such as product categories or customer segments. Temporal data included examples like monthly sales figures or hourly website traffic, which are based on time, while spatial data refers to location-based data.

Step 2: Identify the Relationship Between Variables

Understanding the comparison, distribution, or relationship between the variables is extremely important.

A comparison visual is useful for showing the differences between two or more data points, such as a bar chart or a column chart, while distribution charts can demonstrate how data is distributed, e.g., a histogram or box plot. Lastly, relationship charts, such as scatter plots and bubble charts, can demonstrate how two or more variables are related.

Step 3: Identify the Audience and the Purpose of the Visualization

Next, it’s important to figure out who is going to see the visualization and what the purpose is: what message needs to be communicated through the visualization, complex or straightforward?

If the audience is data-savvy, heat maps or Sankey diagrams can be useful, while simpler pie or bar chart might be more effective for a broader audience. For trends over time, a line or an area chart might be better, while a bar or column graph might be better suited for comparison. For distribution, a histogram or a box plot might be more suitable.If the audience is data-savvy, heat maps or Sankey diagrams can be useful, while simpler pie or bar charts might be more effective for a broader audience. For trends over time, a line or an area chart might be better, while a bar or column graph might be better suited for comparison. For distribution, a histogram or a box plot might be more suitable.

Step 4: Select the Appropriate Visualization Type

Consider the factors above when selecting a chart/graph type. Remember that no visual type fits all, and sometimes multiple visual types may work better to address your message. To find the most appropriate chart type for the dataset, experiment with different chart types.

Different types of graphs

1. Line & Area – Line graphs are useful for illustrating periodic trends, while area graphs show line trends or patterns similar to line graphs but the space between each line filled with a specific colour.

2. Bar & Histogram – Bar graphs are useful for comparing discrete or categorical variables, either horizontal or vertical, while histogram depicts the distribution of numerical data

3. Pictograph – Useful for infographic data presentation using pictures or symbols instead of bars.

4. Scatter plot – Uses dots to represent the correlation between two different variables. Dots form a line closer to each other if there is a stronger relation; otherwise, they demonstrate random dots on the graph.

Different types of charts

1. Pie chart – As the name suggests, the shape of pie splits data into slices as a percentage of a whole.

2. Gantt chart – Displays each task as a bar on the vertical axis and is helpful for monitoring the progress and completion status of each task.

3. Waterfall chart – Uses a colour code to show both the positive and negative impacts of different factors on a value over time, making it useful for financial statements, profit and loss statements, comparing earnings, or budgets versus actuals.

4. Gauge chart – Displays data as a reading on a dial, indicating a specific data point within a specified range.

5. Funnel chart – Illustrates how values progress through different stages, widest at the top and narrowest at the bottom, and is especially useful when tracking the sales process, website traffic, or order fulfilment.