Author: Sophie Lou, Mark Yang
Publisher: Medium
Publication Year: 2020
Summary: The following article is a start-to-finish guide that is good for those uninitiated with the idea of data ethics. It starts by explaining why ethics is relevant to data science. It then briefly explains why we should care about it. It finally references the Oxford-Munich code of conduct and the checklist made by Patil, Mason, and Loukides with some real-life examples. The guide has good wording in saying that we need to understand the underlying human social structures behind our data in order to incorporate human values like equity into our models. We need to check if our models are secure, fair, and transparent in order to avoid some of the disasters that have occurred previously. One example is using personally identifiable information (PII) to detect risks to certain diseases and then marketing certain products more heavily to those groups of people. The article finishes by correctly saying that data ethics is here to stay, so we should “build it in” rather than “bolt it on”.