#22 - The ‘ideal’ class size for online courses does not exist, context will drive the size
Offline or online, facilitating the best learning experience for learners is an ever-present concern for an educator.
One aspect of that is class size.

Class size refers to the number of learners in a class or a session. The debate about the 'ideal' class size is as old as the modern education system itself. There are even legislative provisions regulating Pupil-Teacher Ratio (PTR) and Student-Classroom Ratio (SCR) in Indian schools. These usually range between 20 PTR at primary class levels to 35 PTR for senior classes.
While there is no such legislative provision defining the 'ideal' class size for classes of higher education, the debate has been as intense.
The 'ideal' class size for online courses
The debate about 'ideal' class size in online courses has also received relatively good attention in the last few years. A group of researchers at Stanford University once carried out an experiment engaging about 100,000 student course enrollments. The experiment involved a quasi-random assignment of students to large or small classes. The study found little evidence of the effects of class size on educational outcomes.
Does that mean that class size is irrelevant to discussions on improving learning outcomes in online courses? Not at all!
In my experience at GreyAtom, I have found that large classes work well for foundational courses involving more of factual information acquisition. At the same time, classes with fewer enrollments (40 or less) do better for courses intended to develop higher-order thinking, mastery of complex concepts, and skill development of learners.
Now, there may be a handful of courses that fit strictly in only one of the two categories. Often, a course comprises of foundational learning as well as exercises intended to stimulate higher-order thinking and mastery of complex concepts in learners.
At GreyAtom, we also found that a combination of the different class sizes is better equipped to make use of the best of both worlds.
Our experiments with class size
GreyAtom first began with 25 learners in a physical classroom. This went up to 40 learners with one, and soon, two mentors in a single Hackathon-style learning session. With two mentors, our Hackathons drove the learner’s engagement on fire. Our Hackathons were also designed to be learner-centred. These sessions about their (learners') approach, their questions, and varied perspectives that they needed.
In 2018, when we kicked off online cohorts, we were apprehensive about learners engagement in large groups, so we kept the class size to only 11.
With positive feedback, we again increased the online cohort sizes to range between 40 to 60. At the same time, we experimented with different cohort sizes for different kinds of sessions that we started later on.
Today, we run five different kinds of sessions each with a different cohort size that we deem fit for the respective session.
The cohort size is indicated for each type of session -
A weekly hands-on code along session of 120 mins each consisting of 100+ learners.
A weekly office hour of 60 mins each with about 40+ learners.
A weekly 30 min 1 on 1 session with the career coach.
One-on-one mock interview sessions with mentors.
3 Hackathons and 1 capstone comprising of teams of one mentor and four learners in a total of 88 hours.
End note
With the program structure that we use at GreyAtom (described above), the debate about 'one ideal class size' feels redundant. Our programs make use of large economical cohort sizes as well as provide personal attention in one-to-one sessions.
The key is in the right orchestration!
Do note that large class size does not always mean poor learning outcomes. Even with a cohort size of 250+ learners, some of our mentors drive sessions that set the Zoom chat box on fire🔥 At the same time, a small cohort size is unpractical for foundational lessons.
In other words, a flexible approach to class size in online courses can make use of the best of both worlds. Use them wisely.