AI seen reducing students’ two-year work to six weeks
A recent study by the World Bank Group Education Global Department has found that Nigerian students can make years’ worth of academic progress in just weeks with the help of artificial intelligence (AI).
The study entitled, ‘From Chalkboards to Chatbots: Evaluating the Impact of Generative AI on Learning Outcomes in Nigeria,’ showed that secondary school students in Benin City who used AI chatbots as tutors over six weeks made learning gains equivalent to 1.5 to 2 years of regular schooling.
The pilot involved Microsoft’s Copilot, a chatbot built on GPT-4, and focused on improving students’ English grammar, vocabulary and writing. Students interacted with the AI in 12 after-school sessions, each lasting 90 minutes, with support from teachers.
At the end of the programme, students who used the AI scored about 10 percent higher than those who didn’t. They also outperformed their peers in final exams that included material not covered by the chatbot.
One in six children across the world lives in extreme poverty (or less than $2.15 per day), and according to the Learning Poverty Index, approximately 70 percent of 10-year-olds in low- and middle-income countries cannot read and understand an age-appropriate text.
Read also: Machine learning: Are Nigerian students replacing hard-work with AI
There are about 18.3 million Nigerian children out of school, according to the United Nations Children’s Agency (UNICEF), making the country home to the highest number of out-of-school children in the world.
To assess whether AI could be beneficial, a pilot programme was conducted across nine public schools in Benin City. First-year senior secondary students, approximately 15-year-olds, were divided into treatment and control groups. While 657 students were initially assigned to the AI group and 671 to the control group, only 422 and 337, respectively, completed the final assessment.
Students were grouped based on the number of available computers, with an average of 30 per session, guided by teachers. The final assessment tested their English proficiency based on the national curriculum, as well as their understanding of basic digital concepts and AI.
The assessment was administered in a traditional pencil-and-paper format and was conducted independently by the school, covering the entire term’s content, which extended beyond the six weeks of the after-school programme.
The result showed that students in the AI treatment group had made progress equivalent to nearly two years’ worth of their regular schooling. In end-of-year exams, which covered topics beyond the chatbot’s material, they still did better than their peers.
Despite challenges like internet disruptions and power outages, the intervention still ranks among one of the most cost-effective programmes to improve learning outcomes at a cost of approximately $48 for the six-week pilot and an estimated $124 for a hypothetical four-quarter programme.
Beyond test scores, the study estimated that learning gains could lead to a 14 per cent increase in future wages. The findings also suggested that AI tools could enhance teacher productivity and personalise learning, which is particularly important in a country facing a shortage of over 200,000 teachers, according to the Teachers Registration Council of Nigeria.
The researchers noted that the AI tool worked best as tutors, with teacher guidance and curriculum alignment, rather than merely as shortcuts. However, limitations remain, according to the researchers.
They noted that expanding the programme’s duration and testing it in rural areas to better understand its long-term impact and scalability.
“The rapid development of generative AI presents a unique opportunity to tackle the global learning crisis. By harnessing the responsible use of AI to provide personalised, adaptive learning at scale, governments can take decisive steps toward improving learning outcomes in contexts that have traditionally faced significant educational challenges,” the study added.
