This experiment evaluated the impact of an experimental Spanish-speaking enumerator training on case outcomes, enumerator efficiency, and data quality during the Nonresponse Followup (NRFU) operation for the 2020 Census.

Area census offices (ACOs) were assigned to demographically similar treatment and control groups. The experimental design called for enumerators in treatment ACOs to be assigned and to complete the experimental training module, while enumerators in the control ACOs would not be assigned the experimental training.

The 2020 Census Spanish-Speaking Enumerator Training experiment was motivated in part by almost 600 ethnographic observations of the 2010 Census conducted in English and non-English languages. While both the 2010 Census and 2020 Census included official translations, neither operation included any standard training for bilingual enumerators on when and how to use those translated questionnaires and other non-English materials. 

This experiment’s results found that the experimental training helped Spanish-speaking enumerators to administer the 2020 Census interview in Spanish more accurately and efficiently, and that implementing this training and adapting it for additional non-English languages would benefit future bilingual enumerators.