Start: Oct. 15, 2009
End: Oct. 17, 2009
Place:
The Netherlands
Organizer:
Quantitative Methods in the Social Sciences
Lifecourse analysis is vital to understanding the relationship between life events (e.g. migration, parenthood, partnerships, employment) and long-term outcomes over lifecourse trajectories. The lifecourse consists of repeated events (i.e. multiple episodes) and dynamic interrelated processes with a focus often on the frequency, timing and duration, sequencing, causality and status dependence of events. Topics addressed by the group include:
- Data collection and research design
Life course data are often collected retrospectively and suffer from problems of recall, autobiographic memory, misrepresentation of specific populations, and collecting information on attitudinal or affective states.
- Descriptive methods
Methods are required that can illustrate multistate and multi-episode processes, including: descriptions of entire histories, sequence analysis, analysis of cumulative descriptions over time and graphical methods.
- Life course data and causal modelling
Life course data offer an excellent opportunity to apply causal models, helping to identify: (1) the direction of causality, (2) strengths of reciprocal effects, (3) age, cohort and period effects, (4) historical settings, (5) multiple clocks, eras, and point-in-time events, (6) contextual processes at multiple levels, (7) duration dependence, and (8) variability in state dependencies.
- Multilevel event history models
Individuals may experience events more than once over the observation period. Repeated events data have a hierarchical structure and can be analysed using multilevel models, which can be extended to non-hierarchical structures.
- Dynamic analysis of interdependent events
Most event history analyses focus on the timing of events in a single history or ‘process’. However, outcomes of one process often influence the occurrence of events in another. Multiprocess or simultaneous equation models allow for the examination of interdependencies.
- Using discrete and continuous time models
An important issue is the comparison of discrete and continuous time models and their estimation techniques.
- Cross-national comparisons
Lifecourse data are available in many countries, but cross-national research demands standardized instruments, concepts and compatible data files. The group will develop and test appropriate methodologies for cross-national comparisons.