The survey collected information at school and household level. The objective of the school survey was to collect information:

• on school environmental factors (including the Essential Package elements), 
• school infrastructure and capacity, quality of teachers, availability of textbooks and supplies, and the basic educational performance indicators—school attendance by grade and sex, 
• primary completion rates, 
• rates of entry into secondary school, and 
• average exam scores in basic subjects from a sample of program and control schools. 
• The extent of community participation and the impact of seasonality on attendance were also examined. 
• The objective of the household survey was to provide three critical sources of outcome information to the evaluation:
• livelihood profiles, 
• patterns of educational decision-making (who is sent to school, who stays, etc.) and the stated reasons for these decisions; and 
• the health, nutritional and dietary patterns of the household, especially with regard to children.

The pupil survey will collect anthropometric information from sample students to measure nutritional outcome. The survey was covered two areas namely Char and Coastal under 20 upazilas and 6 districts. The Char and Coastal areas were designed to allow comparison between two strata program and control. A stratify, cluster/school sample was designed in which the schools and households were identified. A total 4 comparison groups were recognized, namely, char program, char control, coastal program and coastal control. A total of 20 clusters/schools per comparison group were selected and total 80 clusters/schools were selected with probability proportional to population. A total 24 households per clusters/school were selected in order to obtain a sufficient number of households per cluster/school.

The survey was awarded by Tango International, USA and funded by World Food Program.