Training to reduce interviewer error: A comparative analysis of two African data collection scenarios

The process of collecting data for epidemiological studies can introduce errors, especially when data are collected in cultural contexts unfamiliar to investigators. Bridging the cultural and language divide in developing countries is essential to guarantee high quality data for studies of environmental exposures and emerging infectious diseases. However, recruiting qualified field staff who speak the local language, are immersed in the local culture, yet understand the demands of scientific research is challenging. When hiring field staff certain qualifications are preferable. However, is higher level education essential? The authors trained, certified, and monitored field teams to conduct separate studies in six cities in Nigeria and seventeen villages in Limpopo Province, South Africa. The training methodology, for both teams, included short lectures, extensive role playing, field practice, and skill assessments. However, the educational backgrounds required of the interviewers were greatly different. In South Africa, interviewers had high school diplomas, while in Nigeria, interviewers had graduate degrees. An analysis of data edit errors from the two studies was conducted to determine whether educational background was related to data quality. Results indicate that despite language and cultural barriers as well as remote location with limited staff education and resources, an effective training program which includes sufficient time for role-playing, field practice and skills assessment can result in the collection of quality data with relatively few errors.