High quality data are critical for providing information on evaluations, but the capacity for high quality data collection varies considerably across developing countries. Our staff work closely with local country data collection partners to ensure the collection of accurate and high quality data for all evaluations. We start by working closely with the implementing partners to ensure that strong survey instruments are designed that capture the details of the interventions and expected outcomes; we then tailor questions to the appropriate country and cultural contexts. We ensure quality data collection using a variety of techniques, depending on the resources and technological constraints of our clients. We have the staff and systems necessary to develop training materials and participate in the training of local data collection staff. We have also developed various data monitoring systems that meet high standards.
In our study in Bihar, we worked closely with local partners to produce a listing of more than 100,000 households to identify women who had a live birth in the past year. We then conducted detailed household interviews with 15,000 women who had recently given birth to obtain information on antenatal care and mother and child health and nutrition outcomes.
In our study examining school dropout prevention projects in four countries, we conducted baseline surveys with thousands of students and teachers to obtain information on attitudes and practices. We are currently extracting school records data for more than 150,000 students in more than 900 schools in these four countries.
In an evaluation of a program intended to strengthen the rule of law, civil society, civic participation, media, and the inspectorate services of the national police in Rwanda, we conducted two rounds of surveys with households nationwide on civic participation issues and perceptions of the police, media, and other relevant institutions.
In the Niger IMAGINE project, we collected data in 189 rural villages throughout the country, which included both household and school surveys, as well as child assessments in math and French. A randomly selected sample of 40 households with school-age children (5 to 12 years old) and up to 3 schools within a 10km radius were surveyed in each village. The process was conducted publicly in each village, and yielded a total of 6,971 households, in which 16,351 school-age children were identified. Based on school registers, we were able to match data collected from 197 schools with 10,858 children in the sample.
For the Niger NECS project, we are working closely with local and international partners as we prepare for the first round of data collection in 204 rural villages in 2013. The data collection process is very similar to that conducted during the Niger-IMAGINE evaluation. We are surveying over 8,000 randomly selected households and over 204 primary schools. We are also conducting assessments in math, French, as well as Haoussa and Zarma (when applicable) of all school-age children in the household sample between the ages of 5 and 14. We are anticipating at least one round of follow-up data collection.