A recent study conducted by economists reveals a significant disparity between the math skills of children engaged in retail jobs and those in academic settings.
In India, many children working in retail demonstrate solid math abilities, allowing them to quickly perform calculations necessary for transactions. However, according to a new study, these children typically struggle with similar types of math problems taught in schools. This disconnect occurs despite many of them still attending school or having completed education through the 7th or 8th grades.
On the other hand, the research also shows that Indian students who are currently in school and do not hold jobs excel at solving academic math problems but often find it challenging to tackle practical problems encountered in a marketplace.
Overall, both the “market kids” and the “school kids” experience difficulties with the type of math the other group excels in, raising concerns about how to enhance math learning for both demographics.
“School students perform worse when transitioning from an abstract to a concrete problem,” explains MIT economist Esther Duflo, co-author of the study. “Conversely, it’s the opposite for market kids.”
Indeed, children who juggle school and work “underperform even though they excel at mental math,” states Abhijit Banerjee, an MIT economist and another co-author. “That was a significant realization for me—that the skills in one area don’t easily translate to the other.”
The paper titled “Children’s arithmetic skills do not transfer between applied and academic math,” is set to be published in Nature. The authors include Banerjee, Ford Professor of Economics at MIT; Swati Bhattacharjee from Ananda Bazar Patrika in Kolkata, India; Raghabendra Chattopadhyay from the Indian Institute of Management in Kolkata; Duflo, Abdul Latif Jameel Professor of Poverty Alleviation and Development Economics at MIT; Alejandro J. Ganimian from New York University; Kailash Rajaha, a PhD candidate in economics at MIT; and Elizabeth S. Spelke, a professor of psychology at Harvard University.
Duflo and Banerjee were awarded the Nobel Prize in Economics in 2019 and are the co-founders of MIT’s Jameel Abdul Lateef Poverty Action Lab (J-PAL), renowned for its contributions to development economics.
Three experiments
The research primarily involved three data-collection exercises, complemented by various experiments. The first revealed that 201 children working in markets in Kolkata possess strong math skills. For example, a researcher acting as a regular shopper would inquire about the price of 800 grams of potatoes priced at 20 rupees per kilogram and then ask about the cost of 1.4 kilograms of onions sold at 15 rupees per kilogram. They would seek the total cost—37 rupees—and give the market worker a 200-rupee note, receiving back 163 rupees. Remarkably, these market children solved such problems correctly 95 to 98 percent of the time by the second attempt.
Nevertheless, when these working children were isolated (with parental consent) and subjected to a standardized Indian national math test, only 32 percent could accurately divide a three-digit number by a single-digit number, and a mere 54 percent could correctly subtract one two-digit number from another twice. Evidently, their practical skills did not translate to academic success.
A second study involving 400 market-working kids in Delhi echoed these findings: while they demonstrated impressive abilities in handling transactions, only about 15 percent who were concurrently enrolled in school achieved average math proficiency.
The second study also posed the opposite inquiry: How do academically successful students perform on market-related math problems? A sample of 200 students from 17 Delhi schools, who do not work in markets, revealed that 96 percent could solve standard problems using paper and pencil, given unlimited time and an opportunity to review their answers. However, when asked to solve similar problems in a simulated “market” context, the success rate plummeted to just 60 percent. It’s important to note that this figure might overstate their capability in a real market environment.
In a third experiment with over 200 kids in Delhi, the researchers compared the performances of both “market” and “school” kids across varied math problems under different settings. While 85 percent of the working children successfully solved a market transaction problem, only 10 percent of the nonworking children answered a similar question correctly when under time constraints and without any aids like paper and pencil. However, when presented with division and subtraction problems that allowed for paper and pencil, 59 percent of nonmarket kids answered correctly, compared to 45 percent of market kids.
To ensure fairness in evaluating both groups, the researchers provided a word problem about a boy purchasing two vegetables. Approximately one-third of the market kids could solve this problem without assistance, while fewer than 1 percent of the school kids succeeded.
What could explain the decline in performance among nonworking students in practical scenarios?
“They learned a formula but didn’t truly understand it,” Banerjee suggests.
In contrast, the market kids appeared to employ specific strategies to navigate retail transactions effectively. For instance, they excelled at rounding: to calculate something like 43 times 11, they might intuitively multiply 43 by 10 and then add 43 for a total of 473. This strategy seems to be common among them.
“The market kids efficiently leverage base 10, which gives them an advantage in base 10 problems,” Duflo remarks. “On the other hand, school kids often do not grasp this concept. The market kids might possess additional strategies that we did not observe.” Conversely, school kids have a stronger command of formal methods for division and subtraction.
Advancing in education
The study’s findings underscore an important issue regarding students’ skills and their academic development. While it’s positive that market-working children are capable of quick calculations, their long-term prospects would likely improve with higher academic achievements like a high school diploma or beyond. Bridging the gap between informal and formal approaches to solving math problems could significantly benefit some Indian children.
The existence of this gap suggests that new teaching methods could be effectively introduced in classrooms.
Banerjee speculates that part of the challenge stems from a classroom approach that suggests only one correct method to solving arithmetic problems. He believes that, following the ideas of co-author Spelke, enabling students to reason their way toward an approximate answer could enhance their comprehension of what is needed to solve these types of problems.
Nonetheless, Duflo emphasizes, “We do not want to place blame on teachers. This situation isn’t their fault; they are bound to adhere to a strict curriculum and specific teaching methods.”
This perspective still leaves the question open regarding what specific changes could be made within the classroom framework. The research team is currently considering new experiments that could address this aspect directly. Nevertheless, the present findings indicate that progress would be beneficial.
“These findings highlight the necessity for educational curricula that connect intuitive and formal mathematics,” the authors conclude in the paper.
Funding for this research was provided in part by the Abdul Latif Jameel Poverty Action Lab’s Post-Primary Education Initiative, the Foundation Blaise Pascal, and the AXA Research Fund.
