Longitudinal surveys are where same group of people are surveyed multiple times over time. This is sometimes also called as cohort or panel surveys. This design is incredibly useful and powerful to gain insights, but they are challenging to implement. We need to be mindful of loss to follow-up, changing migration patterns, and logically fallacies across data from different rounds of survey. Not everyone can do this job carefully. As evaluators, panel surveys are our bread and butter, and we make it look easy!
We have done a panel survey of 500 households which were tracked on bi-weekly basis for an year to measure the cyclical variation in diarrhea and hygiene practices for the World Bank. For IFPRI, we have followed a cohort of 475 women from first trimester of pregnancy to first month after the birth of the child to track diet, nutrition intake, ANC services, weight gain, and birth outcomes. For UC Davis ,we have conducted multiple rounds of panel surveys with 1000+ farmers from 5 districts of Bihar.