Some Methodological Issues in the Analysis of Longitudinal Surveys
Generate an AI Snapshot to get a quick, structured summary of this paper.
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
Abstract
Methodological issues involved in the analysis of longitudinally collected data are examined. Longitudinal studies are ideal for detecting patterns of change. Social scientists ask questions regarding distributional changes and forecasting the evolution of the respondent population along a certain variable continuum. The mathematical framework usually used for such analysis is discrete-time Markov chains. The limitations of this method of analysis to describe social processes are mentioned. Discrimination among alternative structures and use of fragmentary data are discussed. Occupational data is used illustratively.