Testing it and analyzing the results


These interactive videos are strung together with multiple choice quizzes and true/false statement questions, which gauge the user’s conceptual understanding at every level. The algorithms running in the background create a unique and dynamic learning path for each user. This kind of variation in learning speed and style, claims Raveendran, is impossible to achieve without the right tech.

Structuring the learning journey

Byju’s pedagogical design—how they structure the learning journey—is derived from levels in Bloom’s Taxonomy, a popular guide used to design curricula across the globe. Its main purpose is to move students from memorization to application of concepts in real-life situations.

Each concept on the Byju’s platform has its own ‘learning journey’ and is supposed to move a user from Knowledge→ Application→ Evaluation. (Read our previous story on how edtech companies create adaptive learning modules.)

The Byju’s app has collected student data for over three years and has a vast library of “learning deficiencies” or common misconceptions that users face in learning the concept. Raveendran summons a page on a giant, wall-mounted, flat-screen TV with a comprehensive list of these learning gaps. There are so many that one has to squint to read even one.

Every video, says Raveendran, is built keeping these deficiencies (almost 60-70 for each concept) in mind. “Answers from even 10,000 users is enough to estimate 99% of the deficiencies that will crop up during self-study,” says Raveendran. With a total of 33 million registered users, the Byju’s team is likely drowning in such data.

While no one can fault their 1,100-strong content team on the thoroughness and creativity of the self-study modules, how the content is presented and delivered has, so far, never been publicly scrutinized. “Is simply running a complex algorithm on a set of questions better pedagogy? Is it different from computerized GMAT and GRE tests from 15 years ago?

Edtech apps may have a larger question bank or more gradation, but that’s because we have better computing power now. But that’s barely a new pedagogy,” says Meeta Sengupta, a Delhi-based educationist.

For edtech to really work, she adds, it needs to teach concepts in a fundamentally new way, which goes beyond just making interactive videos.

Does Byju’s make that cut?

Raveendran, by his own admission, was an expert at writing stressful exams. Like the Common Admission Test or CAT, which he topped twice, while his friends struggled. Raveendran became an informal CAT-guru overnight.

“I shouldn’t use the word ‘cheating’, but I basically figured out how to crack exam questions,” he says. This is, in all likelihood, why Raveendran’s coaching classes were filling out 1,200-seat auditoriums.

To his credit, Raveendran taught himself maths and science. He learned to speak English by listening to cricket commentary. That others too can be self-taught like him, is a belief baked into the Byju’s platform. And he’s introduced just enough shortcuts and tricks to hook an audience eager to crack the exam code.

“It’s not a flawed goal. Byju’s content and pedagogy will deliver the outcome that students and their parents want from a learning app,” says Lewitt Somarajan, founder of a Pune-based education startup Life-Lab. But he strongly suspects that the outcome is higher marks.

What is the preference?

But marks aren’t Byju’s preferred metric, at least not publicly. The stated objective is to make learning fun. And deeply conceptual.

This is where Somarajan, despite his general approval of the interactive app, exposes a central conflict between the pedagogy and the stated goal. “It’s a hacky way of teaching. The methodology taught in the videos can help you crack an exam question. However, I’m unsure whether this would amount to conceptual learning,” he says.

Sample this, from Somarajan’s review of the “Fractions” module on the app.

The equation is: 1/4 of 100 is 25.

Three different variables here = three different ways of asking the same question.