MOOC: The Data Science Specialization

Image by Alejandro Escamilla, via Unsplash.

Image by Alejandro Escamilla, via Unsplash.

I’ve long wanted to take continuing education classes through a massive open online course (MOOC) provider. I decided that I’d attempt a programming class for for my first venture into the world of MOOCs.

I’ll let you in on a little secret: prior to starting the Coursera Data Science Specialization last autumn, I had never done any programming of any kind. This put me far behind the class curve. Further compounding that mistake was the fact that I chose to start my programming education the hard way, with R, a programming language often used in statistical analysis, particularly in healthcare.

So why R, of all things? Because I’m often handed a clinician’s conclusions regarding a dataset, and I wanted to be able to go back and understand how they came to those conclusions, and whether or not they were accurate. After all, the closer I can get to the raw information, the more effectively I can communicate the outcomes.

The course is run by three professors from the department of biostatistics at Johns Hopkins Bloomburg School of Public Health at Johns Hopkins University. When I registered, the specialization said that no experience was necessary. This was comically untrue. It has since been updated to say that “Some programming experience (in any language) is recommended. We also suggest a working knowledge of mathematics up to algebra (neither calculus or linear algebra are required).” This, in my opinion, is a necessary update. While I had no problems understanding the course concepts, or even the mathematics involved, I struggled endlessly with things that would have been so incredibly basic to anyone who had programmed before, such as How do you load a dataset into R Studio?

Courses included:

  • The Data Scientist’s Toolbox: an introduction to the world of data science
  • R Programming
  • Getting and Cleaning Data
  • Exploratory Data Analysis
  • Reproducible Research
  • Statistical Inference
  • Regression Models
  • Practical Machine Learning
  • Developing Data Products
  • Capstone: a final project intended to pull together the lessons learned in the nine courses

The courses were good, and gave me insight that I wouldn’t have otherwise had. However, there were some challenges with the class structure. For example, I would watch the Week 1 lectures, try to complete the Week 1 project, and find myself thoroughly baffled by what needed to be done. Yet I’d later discover that the Week 3 lectures covered the topic that I needed to know to successfully complete the Week 1 project.

coursera-certificateAfter a few courses, I learned that the only way to approach the classes was to clear my schedule, binge-watch all four weeks of lectures at once, and then proceed with the course work. It made the process dramatically easier.

I also learned that even all these years after school, I still stress about grades. I finished the course with eight 100% scores, a 98.3% in Reproducible Research, and a 97.5% in the capstone; and yes, the capstone grade still rankles a bit.

Would I do it again? Yes, I’ll consider it, though I think I’ll select topics where I have at least a basic level of understanding about the subject. I don’t think it’s an ideal learning environment for someone who is entirely new to the topic.

Have you used MOOCs to expand your knowledge? How did it work for you?

previous article | back to blog index | next article