Evolving online education: Learning Together

Technologies in online learning have been making promises since early 20th century. From radio to present day MOOCs,  multiple new technologies were seen as silver bullet in revolutionising education. These promises were indeed laudable, to make education accessible to everyone, affordable for everyone and more effective. Yet, the same history of unmet expectations has been repeated every time*.

Although a lot has happened over the past decade in terms of growth and expansion of online education, there are many challenges yet to be solved before we can say online learning has truly arrived.

We, at UpGrad, believe the following three challenges are the biggest hurdles to the success of online education.

  • Isolation: Online learning is solitary experience
  • Acceptability: Online learning’s validity in the Job Market
  • Not Engaging: Interplay of distractions and motivations in learning online

The dismally low completion rates and high failure rates are reflection of these challenges [1]. Our goal at UpGrad is to fight these challenges and provide an unmatched learning experience for the working professionals. We believe, with the flexibility of online learning, every working professional can “UpGrade” himself without going back to full-time formal education. Hence, we need to to make online learning mainstream. In this article we would be sharing how are setting the online learning as a ‘Social Experience’ for our students, ie the first of the three mentioned challenges.

In 2000s, internet surfing was mainly an individual experience and on the other hand, in 2016 almost all of internet time is a social experience. We do not feel alone in 2016 version of internet. We believe online education will have much higher engagement and completion rates once it becomes a social experience.

An average student’s e-learning journey is an individual experience making him/her feel isolated. Learners move through the course with limited interactions with both their instructor and classmates. We at UpGrad, are building a new learning platform specifically designed keeping our students and their needs in mind. We believe this would be key to provide the kind of engagement levels the students deserve.

The perception of social presence (or lack of) is a big concern among students and teachers in online learning. Teachers develop courses in isolation and the students take the courses in isolation. Researchers have found strong correlation between the sense of social presence facilitated by the course and student’s perceived learning [3,4,5]. Also there is good re correlation between withdrawal in the course and student’s perceived lack of social interaction and instructor presence [3].

Lack of social interaction in e-learning increases the distraction, since a learner would look for social interaction elsewhere which is usually Twitter, Whatsapp or Facebook. If there is another learner doing the same course in my apartment building, there are limited chances that I would know about him. As learners of online courses, we are oblivious of everyone else’s presence except few replies on the discussion forum. They do form Facebook groups or Whatsapp groups organically, yet the whole online learning format ends up being mostly a solo experience.

Here are the steps we are taking at UpGrad in making our online programs a social experience.:

Profiles :

The first step in ‘social’ e-learning is knowing your classmates. We have learner profiles with education and work history along with other details. Almost everyone is encouraged to fill in their full profile along with a descriptive ‘bio’ section. There are multiple points where learners are nudged to click on the profiles. To improve discoverability of relevant profiles we not only leverage cohort social graph but also use interest graph, performance graph, complementary skills graphs. Combined with other semi random** algorithms like who was recently online, learners bump into each other in interesting ways. In the first month of Data Analytics Program, an average engaged learner viewed peer profiles 110 times.


We have a discussion forum which is contextual and relevant to the course content. We designed it from the ground up to involve learners during the course. They help each other, solve doubts, ask questions, have healthy debates on the forum. Only when there is no consensus a Teaching Assistant gets involved in clearing the doubts. Last 3 months data shows that on a daily basis 50% of the students who are engaged on the platform, also engaged actively on the discussion forum. As most professionals have different backgrounds, sharing of experiences on the forum is much valuable to everyone in the cohort. We have seen learners fall into three buckets on the forum. First is the majority producers, the top 30% of forum content producers are responsible for around 70% of the discussions. The next 50% do the rest of 30% of the content but continue to voice their opinions through upvotes. Rest of the 20% are mere observers. These numbers are highly encouraging and we will be investing more time in finding out how student-student social interactions can help in overall engagement.


Figure 1: Forum topic distribution in one of our programs. Share in the pie is (questions * votes) for each topic. Bigger share of the pie is a proxy for doubts. This lead us to start live sessions on those topics.

Thought Leader AMAs:

Initiating and maintaining engagement on the forum is hard. To set the ball rolling, we started the discussion forum for a new cohort with getting a thought leader to do an AMA on the platform. This way not only the students got a product onboarding experience but also the social expectation was set amongst the peers. In some months, the total content creation during AMAs ended up around 20% of the total cumulative content. We also observed a side benefit of starting the program with thought leader – that learners are more helpful, behave with more civility and show much more respect for each other. This is not very unusual because it is known that a group’s eventual social dynamics are very much impacted by behaviour in initial few days. In a way, it sets the trend which follows upto the end of the course.

We also explored bringing in external motivations to see how does it impact on engagement levels. At times, there were small gifts for top forum contributors and at times we included forum participation in the grading. Our primary observations show that this is to be used cautiously. We are working on establishing a balance between extrinsic and intrinsic motivations for social involvement.

Facebook Groups:

For all cohorts we have Facebook groups for students apart from the platform discussion forum. We were not very confident how much our forum will be used for non academic discussions and generic social sharing. However, we found more and more discussions of generic nature and sharing of articles & blogs also started happening on the forum. Hence facebook group ended up becoming redundant. We are thinking of doing away with a cohort’s facebook group with this expanded use of discussion forum

In summary, having a strong sense of community within the students is a big predictor of learning outcomes. This has been researched multiple times [7, 8]. Our platform’s usage data of course progress and social interaction activity strongly confirms this research. At a time, a learner who is socially engaged on our platform has usually completed 4 times (Figure 2) course materials than a learner who does not participate in social discussions. This high correlation has been repeated across multiple cohorts.


Figure 2: At a snapshot of time course completion rates for learners active on forum vs non active.

Apart from the student isolation e-learning content creators and facilitators tend to work in isolation, reflecting and sharing of best practices is minimal[6]. We at UpGrad are aware of this problem, and making sure different program teams collaborate and share solutions to common problems.

We do not have complete solution to isolation in learning yet. We as a company are focussed on solving this problem, we are testing out few ideas and are willing to share the results with the community. With these changes we are expecting online learning experience to be more engaging, more involving and eventually pushing up the low completion rates of online learning. These are early days of online education and we have miles to go.



[1] http://collegequarterly.ca/2007-vol10-num03-summer/rolfe.html

[2] Online Social Networks as Formal Learning Environments: Learner Experiences and Activities, The International Review of Research in Open and Distributed Learning, Vol 13, No 1 (2012)

[3] Tello SF (2007). An analysis of student persistence in online education. International Journal of Information and Communication Technology Education

[4] Herbert, M. (2006). Staying the course: A study in online student satisfaction and retention. Online Journal of Distance Learning Administration, 9(4).

[5] Morris, T. A. (2009). Anytime/anywhere online learning: Does it remove barriers for adult learners. In T.Kidd (Ed.), Online education and adult learning: New frontiers for teaching practices. Hershey, PA:IGI Global.

[6] Duncan, H & Barnett J (2009) Learning to teach online.

[7] Kaulback, B (2015). Learning Together : Community and Network from the perspective of designers of online learning. (Doctoral dissertation)

[8] Bernard, R. M., Abrami, P. C., Borokhovski, E., Wade, C. A., Tamim, R. M., Surkes, M. A., & Bethel, E. C. (2009). A meta-analysis of three types of interaction treatments in distance education. Review of Educational Research, 79(3), 1243–1289.

The following two tabs change content below.
Ankit Mittal

Latest posts by Ankit Mittal (see all)