When it comes to students’ assessments, most schools have only marks to go by. If a student isn’t faring well in a subject, it’s not easy to understand the reason. The usual suspicions are: “the teacher is not good” or “the syllabus is tough”. The skewed teacher-student ratio only makes appraising each child a difficult task.
Bengaluru-based Digital Aristotle, has developed a platform, AssessEd, which helps schools determine students’ strengths and weaknesses. The solution, which relies on machine learning, gathers data on academic performance, and analyses them to provide actionable insights.
“Our solution augments the existing system in a school, and does not replace it,” says Dev Roy, Founder and MD, Digital Aristotle. “We provide teachers with all the required information, for them to make any corrections, if needed.”
ADVERTISEMENT
Currently mapped to CBSE and ICSE syllabi, the solution covers science, mathematics and social studies. It’s been operational for a little over a year, and 21,000 children in 24 schools are using it, he says.
ADVERTISEMENT
ADVERTISEMENT
Four-stage solution
ADVERTISEMENT
The platform comprises four steps: Question generation, evaluation, reporting, remediation. Teachers can drag and drop questions from a bank that has nearly 2 lakh problems. The questions are categorised under different levels, depending upon how easy it is to solve them. Children answer questions as usual on a paper, which are then scanned and uploaded to the cloud. Teachers correct the answers on a tablet or a computer.
Our solution augments the existing system in a school, and does not replace it - Dev Roy
On each answer sheet, the answers are segregated. This helps teachers to correct the answers to one particular question, of all students in one go. Then, the next question is taken up, and the answers of all students to that question are corrected. This method of evaluating question-wise, instead of the usual answer-script-wise, brings in more consistency and is faster, said Sumod K Mohan, CTO, Digital Aristotle, during a demonstration of the software in their Bengaluru office.
Data analytics brings in loads of insights in the third stage of reporting. Teachers get data on student performance under various heads like topic, chapter, section, ease of answering and more. In the remediation segment, for each wrong answer, students are not only told the right answer, but are also shown from where in the textbook the answer has come from.
Many advantages
Setting the question paper is not just easy, but can also be tailored to required levels of difficulty, by sourcing problems from the appropriate category. Teachers know what percentage of questions are from each category. They get data on which students need more help in specific topics, like fractions, for instance. The data can be sorted in multiple ways, under different parameters.
In mathematics, where answering involves many steps, the Intelligent Tutoring System comes in handy. If a student is stuck at a particular step, he takes a hint with a click. So, depending on the number of hints, teachers know how well the student has understood the concept.
All the data is made available to students and parents. “Usually, at parent-teacher meetings, teachers tell parents that their child is not performing well in a subject, or that he is distracted. But with this data, there is a more meaningful discussion,” says Mohan.
Egalitarian not elitist
The development of the entire solution, for which a global patent has been filed, took about two-and-a-half years, and it involved multiple visits to schools to have a close look at their teaching and evaluation methods, said Roy.
He said since there is so much good material available on the web and in textbooks, they did not want to create something new and force it on students and teachers. “We are using technology to make an egalitarian and not an elitist system,” he said.