D-optimal designs of mean-covariance models for longitudinal data

Area of Science:

• Statistics
• Biostatistics
• Longitudinal Data Analysis

Background:

• Longitudinal data analysis is crucial across many fields.
• Optimal study design requires careful selection of subjects, measurements, and time points.
• Existing criteria often focus solely on mean estimation precision.

Purpose of the Study:

• To develop a novel criterion for optimal longitudinal study design.
• To improve the accuracy of both mean vector and covariance matrix estimation.
• To minimize the total cost of the study design.

Main Methods:

• Utilized mean-covariance models and Bayesian D-optimality.
• Introduced a novel criterion for optimal design selection.
• Developed an algorithm to solve the optimization problem.

Main Results:

• Symmetrization techniques generate improved designs.
• The proposed criterion enhances estimation accuracy for mean and covariance.
• Symmetric and symmetrized designs show superior relative efficiency and parameter estimation.

Conclusions:

• The novel criterion is more effective than previous methods.
• The proposed approach is suitable for maximum likelihood and restricted maximum likelihood estimation.
• Optimal longitudinal study designs can be achieved with improved accuracy and cost-effectiveness.