Response variability in African demographic survey data : a case study of a Nigerian village

This paper is an excerpt from Ken Andoh's doctoral dissertation, "Response Variability in African Demographic Survey Data: A Case Study of a Nigerian Village," written at the Population Studies Center of the University of Pennsylvania. The sources of data for this study are surveys carried out in three villages in Southern Nigeria in December 1973 and Fall, 1974 by Dr. Frank L. Mott (under the auspices of the Human Resources Research Unit, University of Lagos, Nigeria, with the assistance of the Population Council. Vital registration, which is the source of accurate demographic information, is described as inadequate for all tropical Africa. Based on surveys conducted in a Southern Nigerian village in 1973 and 1974, this study seeks to estimate the prevalence and magnitude of misreporting vital events. Survey responses concerning age of respondents, age of children, marital status, duration of residence in the village, educational level, occupation, number of pregnancies, and number of children are presented, and are compared from one survey to the next, revealing great discrepancies between the responses. The degree of correspondence between the surveys was calculated. Certain groups of respondents and certain questions were subject to greater levels of inconsistency than others, with age, number of pregnancies and children, and period of residence exhibiting the highest inconsistency indices.