Fairness in Data Analysis

Hello everyone,

A crucial aspect of being a data analyst is to ensure that the analysis you perform is fair and not biased. Your insights may be right for a particular problem, but they may be biased. As an example, you collected and analyzed evening sales data for an ice cream shop, and you concluded that chocolate ice cream was the most popular item. Although your analysis is accurate, it is based on the fact that you collected and analyzed only evening sales data.

For every given data set you must ask which category of people are excluded. An analyst can be biased for a variety of reasons, for example:

Ways of Collecting Data

Take a Case Study: Your job as an analyst is to study the customer experience of a coffee shop and identify the top-rated coffee. To collect customer feedback, you created a Google form and printed a QR code on the coffee cups. Firstly, customers need internet access to fill out the survey form. There are two telecommunication companies providing internet access in that area. When you conducted the survey one of the companies’ networks was under maintenance, so those using their network were unable to fill out the survey form. Therefore, you have biased results.

Ways of Analyzing Data

Case Study: You must analyze employee records and see who works more efficiently, men or women. The work efficiency of men in the company is higher than the work efficiency of women, so you advise the hiring team to hire men over women. There are 15 men and 2 women employed by the company, so your report is correct but biased. In the case of women, the number of records is not sufficient, so you would have to scale it.

The analysis you're performing may be biased for several reasons. One critical thing to keep in mind is to identify which categories of respondents are excluded.

Thank you