Expression of Interest For Impact Of Ai-Assisted Learning In Telangana: Rct Survey

Expression of Interest for The Impact of AI-Assisted Learning in Telangana: RCT Survey The Office of the Chief Economist for South Asia (DECSA) of the World Bank Group (WBG), in collaboration with the EkStep Foundation and the Government of Telangana, intends to conduct a Randomized Control Trial (RCT) study of the impact of artificial intelligence-assisted learning technology in Indian schools. Foundational literacy and numeracy (FLN) remain critical challenges in India. In response to this challenge, this study explores whether incorporating AI-based tools into the learning process can improve literacy and numeracy for children in primary schools. The study will examine the impact of AXL, a personalized adaptive learning (PAL) tool developed by the EkStep Foundation. AXL is a software that diagnoses childrens initial level and thereafter provides tasks delivered through a computer that are tailored to individual student needs. Students in the treatment groups will receive 1530 sessions (each 30 minutes long) per subject for two subjectsTelugu and Math over two semesters, held every 23 days during school hours. The intervention targets children in grade 3. A randomized controlled trial (RCT) will be conducted with three study arms starting in July 2026 and through to likely February 2027: Control group (C): No additional practice tool, standard education. Practice-only group (T1): Access to basic AXL tool with limited personalization. AI-assisted group (T2): Access to full version of AXL tool augmented with AI and advanced features, including a companion tool for teachers. The impact of the treatments (T1 and T2) will be estimated by comparing the learning outcomes of students in the treatment arms to those in the control arm after the study period concludes. In the control arm (C), students will receive regular classroom instruction. In the treatment arms (T1 and T2), students will be selected by teachers to use the AXL tool as an alternative to regular classroom instruction. These students will be drawn from 600 schools across 12 districts (200 per arm) across the three arms of the study, with an expected average of 15-20 students per school. Each study arm will therefore consist of up to 4,000 students, with a total of up to 12,000 students included in the study. The intervention will be designed and delivered by EkStep and the state education authorities in collaboration with the research team. The objective of this assignment is to collect data to measure the impact of the intervention. The data will be collected in two rounds of student assessment surveys for which enumerators will have to be hired, trained and deployed - conducted in the schools: a baseline round just after the school year commences in July 2026, and an endline following its conclusion (likely February 2027). The assignment also includes conducting and digitizing a short self-administered survey of parents at baseline (disseminated and collected at schools, not at their homes), surveys of school-teachers at baseline and endline, and digitization of other administrative and student background being collected by the research team. The scope of the work involves the following, among other: · Working with EkStep and WBG to obtain necessary access permissions and facilitation from local educational authorities. · Adapting the surveys (student assessment survey, teacher survey, and parent survey) provided by the WBG into the local language. · Recruiting, training and deploying the enumerators that will conduct the student assessment. Field testing and finalizing the student assessment (and teacher survey) through a pilot. · Conducting two survey rounds of student assessment and teacher survey for up to 12000 students. The survey locations will be approximately 600 schools in both rural and urban Telangana. · Data entry, cleaning and timely delivery of data in STATA format. · Qualitative interviews of teachers in 60 schools with instruments that will be provided by WBG team after the endline survey. · Timely delivery of qualitative data in the local language and English. Tender Link : https://wbgeprocure-rfxnow.worldbank.org/rfxnow/public/advertisement/6728/view.html