Table 1.

Covariates for constrained randomization

Covariate	Rationale	Balancing criterion
Socioeconomic status (SES), measured by dichotomizing INSE scores into two categories (> 3 vs ≤3)	SES may predict arboviral infection risk	Each arm within ±5% of overall proportion
Population density	Population density may predict arboviral infection risk	Each arm within ±5% of overall population density
Number of students in grades 1 and 2 combined	Proxy for age of participants to be enrolled in a given cluster. Age is predictive of individual arboviral infection risk.	Each arm 45–55% of total number