Is Online Education the Way Forward?

The world is moving towards MOOCs compared to traditional teaching methodologies due to rising costs of education. Can they be a suitable replacement? (Part 1)

Suhas Motwani
Towards Data Science

--

Earlier this year, I came across an article on the education debt crisis in the US which prompted me to have a look at the alternatives and online education seems to have been a rising star in trying to solve this issue. But is it really the way forward?

Infographic Credits: Niall McCarthy

I find it hard to believe that they can be a suitable replacement. So, is it just a fad? Or is it helping? Let’s find out!

I picked up a dataset that gives MOOC data offered by two leading institutions — MIT and Harvard. (2012–2016)

The Quest:

> What percentage of the participants audited v/s certified?
> Gender, Age demographics
> Possible challenges and interesting insights
> How has total MOOC participation and certification grown over time?
> What percentage of MOOC users earn certificates?

Some Interesting Insights:

The Growth of MOOCs and subsequent student participation

Number of courses as well students have gradually been rising over the years

No doubt, it’s been popular since inception, growing almost 2x every year. Plotting ‘Total participants’ over the last four years shows a decline in enrollments only in the last year — From 1.6M to 1.2M. I would attribute the decline for Harvardx/MITx offerings in the average participants/course and total enrollment for Year 4, due to more competition in the form of MOOC platforms, options, and easier offerings. Harvardx/MITx trends may not be representative of overall trends and hence, this does not imply a decline in online learning.

During these four years, however, 2.4 million unique users participated in one or more MITx or HarvardX open online courses, and 245,000 learner certificates were issued upon successfully completing a course, which leaves us with a completion rate of close to ~10% and an average of 1,554 new, unique participants enrolled per day over four years.

I think if data from other MOOCs appended to this Harvardx/MITx dataset we would likely see a growth, not a decline, in popularity of MOOCs.

The Shift to “Tech”

Computer Science seems to be leading the pack in terms of student participation
  • Computer Science courses are the largest (among, e.g., science, history, health, and other subjects) and route more participants to other disciplinary areas than they receive.
  • We notice the 2014 dip in MIT participants was because of two CS courses by Harvard that became popular. Apart from this, the participation has increased steadily while certification has slowed in the second half. This can be attributed to the cancellation of free certification that was discontinued in early 2016.
  • Further, a look at the course wise breakdown shows us that CS courses have been most popular and CS courses have been the most intensive in
    terms of video hours predominantly due to MITx (the famous Introduction to Programming course!)

Participants in a MOOC “classroom” are heterogeneous in background and intention.

  • A typical course certifies 500 learners — with 7,900 learners accessing some course content after registering, and around 1,500 choosing to explore half or more of a course’s content.
  • Demographic statistics of note include a median learner age of 29 years old, a two-to-one male-to-female ratio (67 percent male, 33 percent female), and significant participation from learners in other countries (71 percent international, 29 percent from the U.S.).

The entire viz can be found here — would love to hear your feedback!

And If Completion rate was to be accounted for..

Credits: http://www.katyjordan.com/MOOCproject.html

HarvardX and MITx recently reported that only 5.5% of people who enroll in one of their open online courses earn a certificate. If taken broadly, these numbers are closer to ~15% but this seems like the only challenge that’s waiting to be solved.

While this initial exploration was definitely exciting, I’d love to dig deeper in Part 2 — Stay tuned!

What’s coming up?

The Quest:

>Understand Behavioral patterns (Cohort v/s Self-paced)

>What I want to do? Talk to folks running various code-schools to understand what’s the secret sauce and the way ahead (so if any of you’ll can connect me to relevant folks, that would be awesome!)

>Uncover data-backed success stories

--

--

Write about Product, Data and Growth| Current: PepsiCo | Ex: Thorogood Associates, Pragmatic Leaders | HMU > www.suhasmotwani.com