Survey research is an integral part of biostatistics because of its focus on methods to facilitate the study of health outcomes and effects in specific human populations. Survey methodology research applies statistical and social science concepts to the development of sample designs, questionnaires, data collection procedures and analysis methods for studies aimed at the general population or targeted population subgroups (e.g., students in schools). The Department of Biostatistics at UNC conducts research in survey methodology through its Survey Research Unit. The Survey Research Unit is committed to improving survey methods through its methodological research and field experimentation. For example, work is underway to gauge the bias and precision implications of methods for dealing with nonresponse in longitudinal health surveys of teenagers. In questionnaire development, SRU researchers are using methods of cognitive psychology to study barriers in interviewing non-English speaking Hispanic women to gather maternal morbidity data. This work will aid the development of interview questions that can be comprehended and answered accurately. Researchers are also engaged in the study of better ways to sample special high-risk population groups, such as migrant farm workers in North Carolina and groups at high risk to contracting AIDS. This survey unit is also engaged in improving methods to obtain small area (e.g., county) estimates from national and statewide samples. In addition to its research mission, the SRU designs samples, evaluates survey questionnaires, and collects data for other university researchers.