Data Science Ethics

DS101x Data Science Ethics is a 4-week survey of ethical issues that arise in data science offered by the University of Michigan through the edX MOOC platform. The course includes 9 modules that begin with a basic overview of ethics and the history of ethics, followed by discussions of data ownership, privacy, anonymity, data validity, algorithmic fairness and society consequences. There are no real prerequisites to take the course, although some familiarity with data science will give you more insight into the material. Grading is based on 9 quizzes (one for each module) that allow unlimited attempts and a written, peer-graded case study. You can earn enough points to pass the course without doing the peer-graded assignment.

Each module is divided into three parts: video lectures, case study lectures and a quiz. The video lecture sections generally consist of about three video segments that run from 4 to 12 minutes each that introduce and discuss major course topics. The lecturer speaks clearly and the slides and video quality are good, but the lecturer's delivery is somewhat monotonous. I recommend increasing the playback speed to keep things moving along at a reasonable pace. The case study lectures look at specific real-life instances of ethical issues presented in the main video lectures. The quizzes are mostly true/false response word problems and you can submit your answers as many times as you want so it is easy to get 100 percent.

Data Science Ethics raises a variety of ethical issues data science practitioners should consider when collecting and using data, but ethics are largely subjective so it can't provide definitive prescriptions about what you should or shouldn't do. It will help you raise relevant ethical questions, but it won't necessarily help you answer them. I found that the case studies were often more relevant and interesting than the main lectures.

Data Science Ethics provides a nice overview of some of the ethical implications of data science and requires a minimal time commitment. Just be aware that the subject matter is subjective so the professor can only really present his take on the issues.

I give Data Science Ethics 3.5 out of 5 stars: Good.