97 Things About Ethics Everyone in Data Science Should Know

Chapter 62. How Can I Know You’re Right?

Majken Sander

Chief Analytics Officer

Of course you strive to be neutral. You more than anyone know that it is essential to avoid skewing results, to stay away from applying any bias, and you find that it is important to let the numbers do the talking.

As a professional on a quest to solve a task, you try to choose the most suitable model and use the best tool available.

Once in a while, someone asks a question like, “How do I choose the best regression model?”

You’re in luck. Years of experience have taught you things such as how to choose the best confidence levels, how to know which values provide the highest success rate in predictive modeling, how to do data cleansing most efficiently, and which records to leave out altogether because of poor data quality.

Data Literacy for Data Users

The people on the receiving end of analytics are seldom aware of what lies behind the data they are looking at. Nor do they have a huge interest in gaining more in-depth knowledge of math, algorithms, and data. Often they find themselves more than happy to trust your craftsmanship and skills. Every once in a while, someone will question the results, but seldom will they question the method.

They need us to take them by the hand and show them the way, maybe even challenge them a bit. Nowadays, a certain amount of data literacy is required for the users ...