In the past year, I've taken several MOOCs offered on both EdX an Coursera platforms. I've been paying money to have my classes "verified" so I can obtain some shiny PDF certificates or Linked-In badges indicating that I actually completed a class. My most recent class is the Columbia University's Machine Learning for Data Science and Analytics course on Edx. In two days I need to decide if I want to shell out $99 to have this class "verified." Unfortunately, this class is not worth it.
Let me explain. To date, I've taken and paid for Georgia Tech's Computational Investing, Columbia U's Financial Engineering and Risk Management Part I, three data science courses from Johns Hopkins, MIT's Introduction to Computer Science using Python and Stanford's Machine Learning. Every single course cost $49 or less. The classes varied in length from 4 weeks to 10. All had moderate to significant amounts of coding or hand-written calculations. Basically, not only did you watch the lecture videos you had to do work to make sure you really understood the material.
The Machine Learning class is really short, 5 weeks long, requires the least amount of assignment work and costs twice as much as the MIT Intro to Comp Sci course. Not only that, the assignments are rather disjointed from the lecture material or requires only superficial understanding. There are quite a few complaints in the classes' discussion group about these gripes. In contrast, the MIT course was 10 weeks long, required answering quiz questions and actual programming assignments. When you completed the MIT class, you knew you learned something .... and it only cost $49!
Sorry Columbia, I think I will keep my $99.