Preparing for a Master’s in Data Science Through Massive Open Online Courses

The era of online education is upon us, and nowhere is this more in evidence than with the proliferation of Massive Open Online Courses: MOOCs.

From their first appearance in 2008, MOOCs have represented an easy way for students with an Internet connection to get the same education from the same professors teaching at some of the finest universities in the world.

With the low cost of digital storage and streaming, many colleges were moving coursework and some class interaction online anyway. Opening those resources up to participants outside the matriculating student body was a trivial step, and one in keeping with the aspirations of many educators.

Stanford University was among the first to open up many of its regular courses to non-students via the Internet. But even before that, MIT’s OpenCourseware initiative had opened the door to distributing previously embargoed teaching materials online for free. Like many other major learning centers, MIT cites the program as a core component of a significant commitment to open, worldwide education.

Eventually, specialty MOOC providers emerged who offered nothing but massive open online courses.

Today, students can find almost any type of course content available online, for free or for pay, with a variety of pedagogies and focuses. Many of these courses pertain to Data Science.

Taking a MOOC may be an alternative or accessory to a conventional degree program. The courses offer an opportunity to pick up specialized knowledge that may not be available locally or in the traditional program a student is enrolled in. MOOCs offer unprecedented opportunity to prepare for graduate studies in a field without going through a traditional bachelor’s program. And some MOOCs now offer certifications, providing a rough analog to a bachelor’s degree.

As such, MOOCs have much to offer students preparing to enter a master’s program in data science even if they do not have the engineering or scientific background commonly recommended.

cMOOCs and xMOOCs

Not all MOOCs are created equal. But there are two broad categories that broadly describe most MOOCs available to data science students today.

cMOOCs Emphasize Connectivity Between Students

A cMOOC-style course may not even have a teacher. Instead, it resembles a massive study group, in which students both learn and help one another learn the material collaboratively.

Unlike xMOOCs, which may be self-paced or allow students to start at any time, the connectivist-style of a cMOOC requires that students start on an established date and progress with their fellow students.

xMOOCs Mirror Traditional Classroom Environments

An xMOOC-style course is often simply an online analogue to a real-world course taught somewhere by a professor. Most MOOCs offered by traditional universities are xMOOCs. In some cases, actual college credit may be offered for completing these courses.

MOOCs vs Bootcamps        

A data science bootcamp is another common method to kickstart entry into a data science master’s degree, but it is a different path than a MOOC. A bootcamp represents an intensive, and usually in-person, education in the field following a scripted pedagogy and hands-on approach.

Educational bootcamps are structured on similar assumptions to the military bootcamps from which they take their name: every person entering them is presumed to start at the lowest common denominator and must be built up from there. Students who happen to have a solid background in statistics already, for example, will still have to study and complete those portions of the bootcamp course to finish the program.

MOOCs, on the other hand, offer a more choose-your-own adventure format. Candidates can select only the courses they need to meet graduate program admissions requirements or that they find most personally relevant. Because pricing is on a per-course basis, the overall cost can be lower than with a bootcamp.

Bootcamps tend to forge friendships, and networking is an explicit part of their offering. Many bootcamps have integrated job placement programs or a career day for graduates. Bootcamps tend toward practical, rather than theoretical, applications of knowledge—the instructors are often practicing data scientists and the coursework usually consists of actual data science projects, which will form a portfolio participants can use on their resume when seeking a position.

MOOCs are closer analogs to a college environment, where concepts and theories are taught in preference to practical applications. A MOOC is more appropriate for candidates who have a good general background for a data science master’s program but merely need to brush up on one or a few specialized concepts. Since each MOOC is a discrete course in a relatively narrow field of study, they can be both more in-depth and less time-consuming than a bootcamp approach.

Where To Find MOOCs

The big players in the MOOC world today are Coursera, Udacity, Khan Academy, and EdX.


Peter Thrum, co-professor of the seminal Stanford CS221, is the founder of Udacity: Introduction to Artificial Intelligence MOOC that signaled the introduction of prestige universities into the MOOC space. The company today prefers to partner with major technology industry businesses rather than colleges for source content and course instructors.

  • Type: For-profit, xMOOC
  • Class delivery: Video lectures with integrated quizzes
  • Class starting restrictions: Open-enrollment
  • Course cost: Mix of free and paid courses; $200/month for paid offerings
  • Grading: Automated grading
  • Accreditation/Certification: Not accredited; Offers “Nanodegree” authenticated certificates
  • Of Particular Interest to Data Scientists: Data Analyst Nanodegree


Coursera was also spawned from Stanford professors but has remained close to its collegiate roots by continuing to offer class content and instructors directly from major university partners. These include Stanford, Princeton, Duke, and Johns Hopkins.

  • Type: For-profit, xMOOC
  • Class delivery: Video lectures, online quizzes and practice exercises
  • Class starting restrictions: A mix of open-enrollment and fixed-date courses
  • Course cost: All courses auditable for free; certification requires $30 – $50 per class
  • Grading: Automated and peer-review grading
  • Accreditation/Certification: Offers “Specialization” completion certificates
  • Of Particular Interest to Data Scientists: Johns Hopkins’ Data Science Specialization

Khan Academy

Khan Academy began life as a series of YouTube videos in which founder Salman Khan doodled out his inventive explanations of math and science concepts for a cousin who needed tutoring. The videos become wildly popular and Khan went on to found Khan Academy as a non-profit educational venture dedicated to producing such content. There is still some debate as to whether or not the organization properly offers MOOCs, but over time it has evolved online exercises and student participation to the point where it closely resembles other MOOC offerings.

  • Type: Non-profit, cMOOC
  • Class delivery: Video lectures, practice exercises, interactive forums
  • Class starting restrictions: Open-enrollment
  • Course cost: Free
  • Grading: Peer-review
  • Accreditation/Certification: None
  • Of Particular Interest to Data Scientists: Probability and Statistics class


EdX evolved out of a cooperative initiative by MIT and Harvard to offer classes online. Today, EdX includes classes from more than 70 other schools, non-profit organizations, and businesses. The organization has also made its learning platform open-source and distributed it freely to other educational institutions. As a result, independent providers around the world now use the EdX platform to teach online classes.

  • Type: Non-profit, xMOOC
  • Class delivery: Video presentations, exercises, online labs, discussion forums
  • Class starting restrictions: A mix of open-enrollment and fixed-date courses
  • Course cost: Free; $25 fee for identify verification for certificates
  • Grading: Automated grading
  • Accreditation/Certification: Certificates offered, some classes eligible for college credit
  • Of Particular Interest to Data Scientists: Data Analysis and Statistics courses

MOOC Aggregators

There are far more MOOC providers today than can be easily listed in one place, which is where MOOC course aggregators come in.

These sites function essentially as search engines for MOOCs from a range of providers, allowing you to find exactly who out there is offering what you want at a price you can afford. They can also help you narrow down enrollment periods for various courses offered only on fixed dates.

Data Science is a relatively new term describing a field that is still being defined. It overlaps with statistics, machine learning, and computer science. And there are many tools and techniques from other fields that are useful to data scientists.

Don’t restrict your searches to “data science” or you will miss many applicable courses. Instead, use these key words when looking for data science courses listed on these aggregator sites:

  • Probability and statistics
  • Computer science
  • Artificial intelligence
  • Machine learning
  • Analytics
  • Big Data
  • Data Mining
  • SQL, R, Python, Java

What To Look For in a Massive Open Online Course

The big decision data science master’s candidates need to make up front when looking at MOOC courses is whether or not to enroll in a complete program, or to just sign up for a handful of unrelated courses to cover knowledge gaps in preparation for a conventional master’s program.

Several MOOC providers offer complete authenticated certificate programs, or genuine college credit from educational partners. Certificates can provide a quick path to employment in the industry. With company’s starved by a shortage of qualified data scientists, a vocational approach is perfectly acceptable to many.

Other MOOCs are offered directly by universities and can provide credits toward their existing degree programs. These classes can serve both as preparation for entering a master’s program and progress toward earning an advanced degree.

Some MOOC providers also offer job placement services (usually for an additional fee) for students completing their courses. If finding a position to use your skills is a priority, consider a provider such as Udacity or Coursera, each of which offer resume listing or placement programs.

Cost and Benefits

Price is almost always a factor and data science MOOCs veer between free and hundreds of dollars each month. Fortunately, it’s easy to use a MOOC search engine to locate courses within your budget range. A MOOC offering job placement services or college credits for completion will generally cost more, but may be of more benefit depending on your goals.


With all the nuts and bolts discussion of course mechanics, it’s easy to lose sight of the most important aspect of taking a MOOC: learning! It’s long been understood that different students learn best in different styles, and MOOCs cater to these differences. The largest difference is between cMOOC and xMOOC style courses; if you are a collaborative learner, a cMOOC might be your best option. If you learn best by watching lectures and doing individual study, xMOOCs might be more your speed.

But there are also differences in presentation that can make a difference. Khan Academy, for example, is known for its straightforward video presentation style, an intimate learning approach that mimics sitting down with an experienced tutor and having a series of informal presentations on the subject—a good fit for visual and auditory learners. But EdX courses have more interactive projects and examples, good for hands-on learners.

Getting the Most Out of Your MOOC

According to one analysis, completion rates for MOOCs hover around 15 percent overall. Certain MOOCs have better or worse completion rates.

MOOC Completion Rate and Assessments

Self-discipline is key. Without the social and institutional pressure to show up for class, perform the work, and take the quizzes, many students—even paying students—fail to complete the courses.

Of course, this is part of the attraction of a MOOC—you are free to get as much or as little out of it as you choose. But most courses are structured so that the curriculum works together as ordered.

Almost all MOOCs allow or encourage students to work with one another via online forums or in-person meet-up groups. This can be a good way to forge a social contract of a sort that might help provide motivation.

It’s also worth your while to check out reviews and comments from prior students for courses you are considering. Some providers have open forums, so you can check out discussions among prior class participants to gauge the level of support and difficulty. Almost all providers also offer course ratings and comments, so you can explicitly see how previous students found the class.

For reviews without any potential bias, independent websites also include class reviews:

Back to Top