Studying developmental processes in accelerated cohort-sequential designs with discrete- and continuous-time latent...

Area of Science:

• Developmental psychology
• Quantitative psychology
• Methodology

Background:

• Studying long-term psychological development is complex.
• Accelerated longitudinal designs (ALD) offer a time-efficient approach by using multiple cohorts.
• ALD data requires advanced statistical modeling to capture developmental trajectories.

Purpose of the Study:

• To evaluate discrete- and continuous-time latent change score models for analyzing ALD data.
• To assess model performance under varying sampling conditions and cohort convergence.
• To examine the recovery of exponential developmental trajectories, such as cognitive development.

Main Methods:

• Monte Carlo simulation study.
• Comparison of discrete-time and continuous-time latent change score models.
• Analysis of data from simulated accelerated longitudinal designs with varying cohort characteristics.

Main Results:

• Accelerated longitudinal designs are appropriate for studying developmental processes across different sampling conditions.
• Both discrete- and continuous-time models accurately recover parameters when cohorts are from the same population.
• Discrete-time models show bias with inconsistent time lags; both models exhibit bias when cohorts lack convergence.

Conclusions:

• Accelerated longitudinal designs are viable for developmental research.
• Continuous-time models are preferable for analyzing ALD data, particularly when cohort convergence is questionable.
• Recommendations are provided for implementing ALD and utilizing continuous-time models in developmental research.