Analysis of longitudinal data with unequally spaced observations and time-dependent correlated errors

Area of Science:

• Statistics
• Biostatistics
• Longitudinal Data Analysis

Background:

• Repeated measurements data often exhibit complex correlation structures.
• Standard models may not adequately capture time-dependent variations within subjects.
• Understanding audiologic performance changes requires robust statistical approaches.

Purpose of the Study:

• To propose a novel linear model for repeated measurements.
• To incorporate a time scale transformation for flexible covariance structures.
• To apply the model to real-world audiologic performance data.

Main Methods:

• Development of a linear model with a transformed time scale in the correlation structure.
• Utilizing restricted maximum likelihood (REML) for parameter estimation.
• Application to simulated and speech recognition data from the Iowa Cochlear Implant Project.

Main Results:

• The proposed model accommodates nonstationary covariance structures, with stationarity as a special case.
• Demonstrated application to audiologic performance data, showing growth curves.
• Provided estimates of standard errors for predictions at specific time points.

Conclusions:

• The enhanced linear model offers a flexible framework for analyzing longitudinal data with time-varying correlations.
• The methodology is effective for modeling audiologic performance trajectories.
• The approach provides valuable insights into predicting outcomes in clinical studies.