Advances in longitudinal research methodology

Issues in longitudinal research in geropsychology are identified. Advances in longitudinal research methodology are illustrated by summarising three symposium contributions. Tjorborn Svensson discussed whether biographical retrospective data can be conceptualised as a longitudinal inquiry, and what the kind of design would be that one would want to apply given a positive answer to the question. Scott Hofer treated the topic of subject attrition in longitudinal studies within the more general framework of contemporary missing data analysis. Christoph Rott addressed the issue of variability within and across individuals. Using the example of the Bonn Longitudinal Study (BOLSA), he showed how model misspecifications can circumvent limitations of the designs using an age-cohort-period model. He also incorporated adjustments for attrition and substitution of an event definition for calendar time.