The Nutrition-sensitive Aspect of the "Development of Sustainable School Feeding” Project in Armenia
How to cite
Alderman, Harold; Knauer, Heather. 2018. "The Nutrition-sensitive Aspect of the "Development of Sustainable School Feeding” Project in Armenia." AEA RCT Registry. AEARCTR-0003581. https://doi.org/10.1257/rct.3581-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
The study will assess the impact of a preschool snack added to the current school meal program in Armenia. The assessment will be based on a randomized controlled trial [RCT] in three provinces with relatively high poverty levels and high rates of under 5 anemia. This will allow a causal assessment of this component of the school feeding project.
The initial round of the RCT will focus on assessing the fluid intelligence of 0 grade students at midmorning. The same group of children (around 1200 students) will again be tested near the end of the school year (Spring 2019) to assess crystallized intelligence, that is, tests that are designed to assess school readiness.
Intervention
Schools will be provided funds for a snack proportional to preschool enrollment along with a suggested list of foods items for snacks. Cheese, eggs and similar products are recommended; sugary snacks are discouraged. The snacks will be provided daily around mid-morning. There are 50 schools pairwise randomized to the treatment and 50 control schools.
Experimental design
All schools in three provinces selected on poverty were visited to ascertain if any provided a snack through community organizations. These few schools were not listed in the sample frame, nor were schools with less than 6 students. All other schools in the provinces were paired on the basis of location and size and pairwise randomization was used to selected treatment schools for a sample of 50 treatment schools and 50 control schools. In each school a random sample of 12 children in grade 0 will be selected to testing. In the event that the school has less than 12 students, all children in the grade will be sampled. These children will be tested in both rounds. The children who are not selected for the testing will still be included in similar games and will also receive the same ‘stickers’ as the others. This will reduce any sense of stigma. First graders were similarly selected randomly to be tested in the beginning first round of data collection
Assuming an inter-cluster correlation [ICC] of 0.12, the sample is powered to detect a significant difference in effect size of 0.25 with p<0.05. While it may be desirable to seek a smaller MDES, it is likely impractical given the number of schools available in the provinces and the size of the student body in many schools. The assumed ICC was based on a review of tests for preschool children, mainly from Latin America; no similar information is available for Armenia nor is there an international repository of such correlations
Design details
Cited by
No citation data available for this registration yet.