Experimental Design
We conduct a three-arm randomized controlled trial to estimate the impacts of remote, phone-based interventions on children’s learning and empowerment outcomes in Afghanistan. The study sample consists of 2,399 households, each contributing one target child aged 7–16 and their primary caregiver. The sample was drawn from a household listing exercise covering approximately 10,000 households across 26 provinces and prioritizes girls of the age group with limited access to education.
Following completion of a baseline phone survey and eligibility verification, households were randomly assigned at the household level to one of three arms: (i) ConnectEd Tutoring, (ii) Structured Conversation, or (iii) a control group receiving no intervention. Randomization was stratified on baseline learning level (above/below median), baseline empowerment (above/below median), schooling status (in school/out of school), and child’s primary language (Dari or Pashto), yielding 16 strata, with an additional stratum for boys. Within each stratum, assignment was approximately evenly split across the three arms.
The ConnectEd Tutoring arm delivers weekly one-on-one phone-based instruction in foundational literacy followed by numeracy, using a teaching-at-the-right-level approach, over an eight-week period. The Structured Conversation arm delivers weekly one-on-one phone calls of equal length with a female teacher, focused on guided conversation topics and explicitly excluding academic instruction. The control group receives no calls during the study period. All interventions are delivered by trained teachers and last approximately 25–30 minutes per week.
Primary outcomes are measured via endline phone surveys conducted after the eight-week intervention period and include assessments of foundational literacy and numeracy and adjacent cognitive skills not taught during the tutoring sessions, measures of empowerment and agency, and measures of well being and gender attitudes.