The road is filled with the chatter of children, and Ha—No, let’s call the student ‘X’, and I’ll reveal why later—X joins the others as they pass through the gates of their school. X goes straight to the grade 9 classroom after the morning assembly. The first class is Mathematics. X has been at this school for only a month, and is trying to catch up with the others. As she does every day, the teacher writes a problem on the board and turns to the class to call on a student to help her solve the problem. X avoids her eyes and looks down, trying to look simultaneously calm and inconspicuous. The teacher calls out another student’s name, and X lets out a sigh of relief. X tries to concentrate on what is happening on the blackboard, but after the third step of the calculations, fails to understand how the numbers are transforming. X wonders, “What did they learn in the previous classes? I can’t seem to understand this… Oh no, the first problem’s done and it’s time for the second one now. Don’t look at the teacher, stay cool!”
The teacher’s eyes fall on X’s bent head and she thinks to herself, “There’s X, trying to be invisible again. I know X’s family has moved to four different towns in the past three years, and X is behind on many topics. I’ll give some extra time to X, Y and Z this week. They all really need it. Oh wait, not this week, I need to make the arrangements for the scholarship exam. I’ll do it next week. I can’t ignore the other students right now – they need to prepare for their board exams next year after all. And our school needs to maintain its passing percentages.”
The bell rings and it’s time for the students to head to the computer lab. X sits at the computer and logs into Mindspark, a learning software. The questions coming up on the screen are challenging and engaging, but not daunting, to X because they are actually from a grade 7 concept. X needs to catch up on various concepts, and Mindspark allows X to do this. But how does it identify what X needs?
Mindspark is a personalized and adaptive program that uses a multi-pronged approach to solve this problem.
When a student logs into Mindspark for the very first time in a subject, they are given a diagnostic test that determines the overall level at which the student lies on the spectrum. This means that even if a student is in grade 7, they could be given grade 3 content, depending on the level determined by the diagnostic test.
In this way, each student starts their journey at the level best suited to them.
Checking for conceptual learning:
Questions in Mindspark are specially designed to test understanding and to help students clear their misconceptions. When a student answers a specific question or combination of questions incorrectly, the system diagnoses the child’s misconceptions / weak areas. The child may be further provided with a simple or detailed explanation, or be redirected to questions that strengthen basic understanding.
Thus, the system does not allow a student to move up to higher levels without a strong understanding of the basic concepts. This helps students below the average class level to come up to their class level. Similarly, it provides challenging problems to high performing students allowing them to stay engaged, and enabling them to learn more. These decisions are driven by the adaptive logic which is designed to improve over time with increased student usage.
Pinpointing weak areas:
Mindspark examines patterns of errors to target “differentiated remedial instruction.” For instance, if a student makes a mistake on which decimal is bigger (3.27 or 3.3), it may be due to “whole number thinking” (27 is bigger than 3) whereas if they make the same mistake with 3.27 or 3.18, it could be due to “reverse order thinking” (comparing 81 to 72 because the “hundredth place” should be bigger than the “tenth place”). A good teacher may catch this if most of the class is making the mistake, but the likelihood is low if only a few students are making the error.
Thus, to tackle the need for specific and personalized learning paths for each student, the adaptive logic in Mindspark uses a student’s performance data as well as the conceptual framework for different topics to deliver relevant content.
An example of a personalized learning path for a student is given below:
Based on research at the University of Melbourne by Kaye Stacey, et al.
Remediating learning gaps:
In Mindspark, when a student gets 25% or more questions incorrect in a learning unit, the program allows the student to repeat the learning unit. If the student got the questions incorrect due to a lack of understanding of a fundamental concept, then the adaptive logic takes the student to the immediate previous learning unit. If the student got the questions incorrect due to a misconception, then a remedial module is given to the student. Remedial modules are designed to resolve the student’s misconception based on cognitive dissonance theory. In the cognitive dissonance method, the system conflicts the student’s prior understanding (misconception), then gives the correct explanation to resolve the misconception. 1
Does this approach work?
The Abdul Latif Jameel Poverty Action Lab (J-PAL), a Research Centre at Massachusetts Institute of Technology led by Professor Karthik Muralidharan (co-chair of education), conducted a pilot randomized evaluation of the Mindspark program, deployed through stand-alone centres in Delhi as well as another subsequent evaluation in Government schools of Rajasthan.
Some of the key findings were:
- The program was equally effective for students at all levels of the achievement distribution. Treatment effects in the impact evaluation did not vary significantly by level of initial achievement, gender or wealth. Thus, the intervention was equally effective in teaching all students.
- However, the relative impact of the program was much greater for low-achieving students, who were making no progress in school. While the absolute impact of Mindspark was similar at all parts of the initial test score distribution, the relative impact was much greater for weaker students because the “business as usual” rate of progress in the control group was close to zero for students in the lower third of the initial test score distribution.
- The program had large effects on student achievement in Math and Hindi. The usage of the program led to doubling of student test scores in math and Hindi over a 4-5-month period. There was a linear correlation between usage and gain.
But why was the student named X?
But why was the student named X?
Student X’s anonymity at the start of this article was to highlight the fact that a name is often linked to a variety of characteristics that lead to potential discrimination within the classroom – characteristics like gender, religion, caste, etc. Such experiences compound the problems faced by students on their learning journey.
A learning software, like Mindspark, does not consider such characteristics, thereby empowering the students to focus on their journey. The only aspects that Mindspark focuses on are whether the student is learning, and what the student is struggling with. This becomes especially important when a student is not able to receive the required support from the home environment in order to overcome possible challenges at school.
Mindspark is built on a simple principle and vision – helping each student learn with understanding – and its adaptive logic creates personalized paths to help each student realize their potential. In doing so, it also helps the teacher deal with heterogeneity in the learning levels of students within a single classroom.
1 Rajendran, Ramkumar & Muralidharan, Aarthi. (2013). Impact of Mindspark’s Adaptive Logic on Student Learning. Proceedings – 2013 IEEE 5th International Conference on Technology for Education, T4E 2013. 119-122. 10.1109/T4E.2013.36.
Muralidharan, Karthik, Abhijeet Singh, and Alejandro J. Ganimian. 2019. “Disrupting Education? Experimental Evidence on Technology-Aided Instruction in India.” American Economic Review, 109 (4): 1426-60. https://www.aeaweb.org/articles?id=10.1257/aer.20171112
Vincy Davis, Anupriya Singh et al. (2019) Report on Time Allocation and Work Perceptions of Teachers. Accountability Initiative, Centre for Policy Research, New Delhi. https://accountabilityindia.in/wp-content/uploads/2019/06/REPORT-ON-DELHI%E2%80%99S-GOVERNMENT-SCHOOL-TEACHERS-.pdf