Survey data has never been as indispensable as it is for XXI century society, especially because this type of data is the main source of information when regarding demographic and social characteristics of the population, economic activity, lifestyle patterns, and public opinion (Barnett, 1991). Longitudinal household surveys often use a complex sampling design to select the sample to be followed up over time. It is well known that complex sampling schemes may inflate the variances of estimators, especially as a result of clustering. This book presents evidences that design effects for longitudinal analyses can be greater than for corresponding cross-sectional analyses, implying that more caution is required before ignoring the complex design in standard error estimation. Furthermore, methods for the analysis of longitudinal data collected under complex sampling designs are discussed, with emphasis given to covariance structure modelling techniques. Methodology presented in this book has wide application in the Social Sciences.

Данное издание не является оригинальным. Книга печатается по технологии принт-он-деманд после получения заказа.