Reducing Bias

A bias refers to our tendency to favor specific perspectives or opinions over others, resulting in skewed conclusions that may not accurately reflect reality. In data storytelling, a bias can occur when the way data is presented tilts toward a specific interpretation of the information. One example of data visualization bias is inappropriate graphs or chart scales.

Consider the following two techniques to reduce biases in a chart:

• Checking the data

• Checking the design elements

Checking data

Cherry-picking data is commonly used to support a particular narrative. It involves selectively choosing data that supports our own argument, while omitting other relevant information that may provide a complete picture of the situation. This technique can lead to biased and inaccurate conclusions.

Let’s suppose a researcher wants to show that the sales of Product A are increasing over time and creates a line chart to visualize the data. However, instead of plotting all the data points, the researcher only includes the data points for the months where the sales of Product A were the highest.