Can Learning Analytics Be the Next Big EdTech Breakthrough?
Share
Every answer, pause or retry on a digital learning platform leaves a trail. It can show more than whether a student is right or wrong: where she is confident, where she hesitates and which mistakes recur.
Analysing such records is known as ‘analytics’. Artificial intelligence (AI) can spot patterns and adapt practice to each learner. The National Education Policy (NEP) 2020 recognised this potential, noting that AI-based software could help students track their growth through learning data. The bigger question for teachers is whether it can help them understand each learner better.
The learning gap that develops quietly
Learning gaps rarely appear all at once. A student misses one concept, manages the next lesson and moves on; until the gap surfaces in a test months later.
That is difficult to catch at scale. India’s 98 lakh teachers serve 24.8 crore students, according to the Economic Survey 2024-25. The PARAKH Rashtriya Sarvekshan, which assessed 21.15 lakh students, found that only 53 percent of class VI students were comfortable with arithmetic operations and multiplication tables up to 10; the national average in class VI mathematics was 46 percent.
The numbers show the scale of the problem. They do not show where an individual learner went off track; or what might bring her back. That requires a closer look at learning as it happens.
Analytics fills part of that gap. It might show that a student solves addition quickly but slows down when carrying over, or that repeated errors in fractions point to a weak grasp of place value. These are not complete explanations, but signals a teacher might otherwise miss.
This closer view becomes more useful when assessment happens throughout the learning process. NEP 2020 calls for regular, formative and competency-based assessment. The AI and computational thinking curriculum for classes III-VIII similarly includes continuous assessment and teacher observation. Together, they point to a shift from asking what a student scored to asking what the evidence says about how she is learning.
What the data cannot explain
If a student’s scores fall over two weeks, the data can flag the change. It cannot tell the teacher whether she has misunderstood a concept, lost confidence or simply found the work unchallenging. A 2023 US Department of Education report similarly stressed the role of people in recognising patterns and deciding what they mean.
That is the line AI cannot cross on its own. The data can raise the question. The teacher has to answer it.
AI’s value is greatest when it helps a teacher see something sooner.
Consider a small coding class, where students are learning loops and building their first programs. The teacher introduces the concept and students build a short program. A quiz and worksheet reveal the gap. Two students keep making the same error. Between classes, an AI tutor gives them extra practice and hints. By the next session, the teacher can see whether they still need help. If they do, she can explain the idea differently, pair them to work through it together or give a confident learner a more challenging task.
The technology has not made the decision. It has shortened the distance between a difficulty appearing and a teacher responding.
Measuring what matters
That raises another question: what gets lost when learning becomes data?
A platform can count speed and accuracy. It is much harder to capture curiosity, persistence or the satisfaction of finally solving a difficult problem. A student building a game, an app or an AI-powered tool may reveal more about how she thinks through the project than any score can show.
The same care is needed with the data itself. The Digital Personal Data Protection Rules require verifiable parental consent before a child’s personal data is processed, subject to specified exemptions. Parents should also be able to understand what the data says about their child’s progress. Regular reports and parent-teacher conversations can turn analytics into a shared understanding rather than another number on a dashboard.
From knowing more to teaching better
A national assessment can show the scale of a learning problem. Classroom data can show where an individual learner may be struggling. A teacher can then bring the missing context: confidence, curiosity, circumstances and the way that particular child learns.
That is the real opportunity for AI in education: not to take the teacher out of the loop, but to give the teacher a better view of every learner in it.
Views expressed are personal
The author, Vivek Prakash, is CEO at Codingal
