Analysis of a within-subject design with covariates

Area of Science:

• Statistics
• Biostatistics
• Data Analysis

Background:

• Analyzing incomplete data in within-subject designs presents statistical challenges.
• Monotone missing data patterns require specialized analytical approaches.
• Baseline measurements are crucial for understanding treatment effects.

Purpose of the Study:

• To develop a method for analyzing incomplete data with monotone patterns.
• To provide adjusted means and dispersion matrices for period-by-sequence cells.
• To propose statistical tests for hypothesis evaluation in such designs.

Main Methods:

• Assumed multivariate normality with antedependence structure for measurements.
• Employed methods to obtain adjusted means and dispersion matrices.
• Developed large sample tests for hypothesis testing.

Main Results:

• The methodology successfully analyzes incomplete data with monotone patterns.
• Adjusted means and dispersion matrices were derived.
• Large sample tests were proposed for hypothesis evaluation.

Conclusions:

• The proposed method offers a robust approach to analyzing incomplete within-subject data.
• The methodology is applicable to scenarios with monotone missing data patterns.
• The numerical example demonstrates the practical utility of the approach.